{"generated_at":"2026-10-16T14:01:47-05:00","count":241,"source":"https://chatppg.com/research","items":[{"slug":"para-handcycling-sci-healer-multimodal","name":"A Longitudinal and Multimodal physiological signals dataset of Para handcycling training and race for SCI Rehabilitation","tagline":null,"summary":"This dataset follows one para-athlete with a spinal cord injury through five months of training that ended with the first hand-bike ride across Saudi Arabia. He trained with the portable VIK8 system using adaptive functional electrical stimulation kinesitherapy (AFESK) from VIKTOR S.r.l.","signals":["ecg","resp","temp","spo2","accel"],"subjects":null,"license":"CC BY 4.0","gated":null,"updated":"2026-10-14","url":"/research/datasets/para-handcycling-sci-healer-multimodal","links":[{"href":"https://doi.org/10.5281/zenodo.23168228","label":"DOI","ariaLabel":"DOI for A Longitudinal and Multimodal physiological signals dataset of Para handcycling training and race for SCI Rehabilitation (opens in new tab)"}]},{"slug":"multimodal-nback-music","name":"A Multimodal Dataset for Investigating Working Memory in Presence of Music","tagline":null,"summary":"This dataset accompanies a study of whether music can regulate cognitive arousal and performance during a working memory task. Participants did 1-back and 3-back tasks with calming music in the first session and exciting music in the second, with 22 trials per block. Physiological signals and behavioural data were recorded throughout.","signals":["ppg","ecg","eda","eeg","resp","skin-temp","accel"],"subjects":null,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/multimodal-nback-music","links":[{"href":"https://physionet.org/content/multimodal-nback-music/1.0.0/","label":"Download","ariaLabel":"Download for A Multimodal Dataset for Investigating Working Memory in Presence of Music (opens in new tab)"},{"href":"https://doi.org/10.13026/6vh4-dk68","label":"DOI","ariaLabel":"DOI for A Multimodal Dataset for Investigating Working Memory in Presence of Music (opens in new tab)"}]},{"slug":"aauwss","name":"AAUWSS","tagline":"Empatica E4 + PSG overnight sleep staging (n=13)","summary":"AAUWSS (Djanian / Nielsen / Nielsen / Bruun, Aalborg University + Aarhus University Hospital Respiration Center West). Collected Mar 24–Apr 4 2023. Release includes raw Empatica CSVs, PSG EDFs, AASM annotations, and alignedsleepdataset pickles (30 s epochs: ECG 30×200 Hz, PPG 30×64 Hz + labels).","signals":["ppg","ecg","accel","skin-temp","eda","eeg","eog","psg"],"subjects":13,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/aauwss","links":[{"href":"https://zenodo.org/records/16919071","label":"Download","ariaLabel":"Download for AAUWSS (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.16919071","label":"DOI","ariaLabel":"DOI for AAUWSS (opens in new tab)"}]},{"slug":"adarp-aud-stress","name":"ADARP (AUD EMA stress, E4)","tagline":null,"summary":"ADARP comes from a pilot study of people diagnosed with alcohol use disorder, built to see how their daily experiences line up with physiological signs of stress. Participants answered web surveys about emotions, cravings and stress four times a day for up to 14 days while wearing an Empatica E4 wristband that recorded heart rate, electrodermal activity, skin temperature and movement. Structured interviews with a timeline follow-back calendar were used to log daily drinking and check the self-reported and physiological stress markers.","signals":["ppg","accel","skin-temp","eda"],"subjects":11,"license":"CC BY 4.0","gated":null,"updated":"2026-10-14","url":"/research/datasets/adarp-aud-stress","links":[{"href":"https://zenodo.org/records/6640290","label":"Download","ariaLabel":"Download for ADARP (AUD EMA stress, E4) (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.6640290","label":"DOI","ariaLabel":"DOI for ADARP (AUD EMA stress, E4) (opens in new tab)"}]},{"slug":"aspen-adelaide-wearable-psg","name":"Adelaide ASPEN wearable-PSG garment","tagline":"1-lead ECG + respiration + body position/actigraphy + skin temp (7 OSA patients, 63 h, sleep lab)","summary":"This dataset comes from ASPEN, a wearable sensor that records respiration, single-lead ECG, body position and activity, and skin temperature for sleep studies. Seven men with diagnosed sleep apnoea (mean age 55.7, mean BMI 29.9) wore it during research polysomnography at the Flinders University Sleep Lab, giving 62.94 hours of data. Each recording was checked against the standard polysomnogram taken at the same time.","signals":["ecg","resp","skin-temp"],"subjects":7,"license":"CC BY 4.0","gated":null,"updated":"2026-10-14","url":"/research/datasets/aspen-adelaide-wearable-psg","links":[{"href":"https://doi.org/10.6084/m9.figshare.33399838.v1","label":"Download","ariaLabel":"Download for Adelaide ASPEN wearable-PSG garment (opens in new tab)"}]},{"slug":"afpdb","name":"AFPDB","tagline":"PAF Prediction Challenge Database (CinC 2001)","summary":"PhysioNet afpdb 1.0.0 (Moody / Goldberger / McClennen / Swiryn). Learning records n (no documented AF) and p (PAF subjects, with 5-min continuation c showing PAF onset on even-numbered precede records). Test records t01–t100 without continuation files.","signals":["ecg"],"subjects":98,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/afpdb","links":[{"href":"https://physionet.org/content/afpdb/1.0.0/","label":"Download","ariaLabel":"Download for AFPDB (opens in new tab)"}]},{"slug":"ai-readi","name":"AI-READI","tagline":"Flagship type 2 diabetes dataset (FAIRhub): Garmin Vivosmart 5 wrist HR/SpO2/stress + Dexcom G6 CGM + clinical, ECG and retinal imaging","summary":"Artificial Intelligence Ready and Exploratory Atlas for Diabetes Insights. Versioned releases on FAIRhub: v1.0.0 (2024-05-03, DOI 10.60775/fairhub.1), v2.0.0 (2024-11-08, DOI 10.60775/fairhub.2; 1,067 participants, data collected 2023-07-19 → 2024-07-31), v3.0.0 (2025-11-17, DOI 10.60775/fairhub.3; 2,280 participants per the v3 healthsheet). Annual releases planned.","signals":["hr","spo2","accel","cgm","ecg"],"subjects":2280,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/ai-readi","links":[{"href":"https://fairhub.io/datasets/3","label":"Download","ariaLabel":"Download for AI-READI (opens in new tab)"},{"href":"https://doi.org/10.60775/fairhub.3","label":"DOI","ariaLabel":"DOI for AI-READI (opens in new tab)"}]},{"slug":"airwave-health-ppg-pwv","name":"Airwave Health Monitoring Study","tagline":"clinic finger PPG + Vicorder PWV (Imperial / Elliott)","summary":"The Airwave Health Monitoring Study was set up to check for possible health risks from TETRA, the digital radio system used by British police and emergency services since 2001, and later widened to the general health of that workforce. Enrolment started in 2004, either by questionnaire or by a local health screening. By the end of 2012 it had recruited 42,112 participants, and 35,199 of them (83.6%) attended the screening.","signals":["ppg"],"subjects":null,"license":"CC BY-NC-ND 3.0","gated":null,"updated":"2026-10-15","url":"/research/datasets/airwave-health-ppg-pwv","links":[{"href":"https://police-health.org.uk/","label":"Download","ariaLabel":"Download for Airwave Health Monitoring Study (opens in new tab)"},{"href":"https://doi.org/10.1016/j.envres.2014.07.025","label":"DOI","ariaLabel":"DOI for Airwave Health Monitoring Study (opens in new tab)"}]},{"slug":"alameda-pd","name":"alameda-pd","tagline":null,"summary":"The ALAMEDA dataset holds raw wrist accelerometer data from 11 people with Parkinson's disease wearing GENEActiv bracelets. Patients were recruited at a Parkinson's clinic at Eginitio Hospital in Athens and followed for a year, with four intensive one-week recording periods, each before a three-monthly clinic re-evaluation. It is meant for research linking upper-limb activity patterns to clinical status and for building digital biomarkers.","signals":["accel"],"subjects":11,"license":"CC BY 4.0","gated":null,"updated":"2026-10-14","url":"/research/datasets/alameda-pd","links":[{"href":"https://doi.org/10.5281/zenodo.15769959","label":"DOI","ariaLabel":"DOI for alameda-pd (opens in new tab)"}]},{"slug":"aquaset","name":"AQUASET","tagline":"affective QoE streaming (Empatica + Muse + Movesense)","summary":"AQUASET (Affective Physiological QUAlity-of-Experience Streaming service dataSET) combines streaming QoE parameters, subjective assessments, physiological signals, and facial-expression features. Zenodo 22741195, DOI 10.5281/zenodo.22741195, license CC BY 4.0.","signals":["ppg","ecg","accel","skin-temp","eda","eeg","ibi"],"subjects":30,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/aquaset","links":[{"href":"https://zenodo.org/records/22741195","label":"Download","ariaLabel":"Download for AQUASET (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.22741195","label":"DOI","ariaLabel":"DOI for AQUASET (opens in new tab)"}]},{"slug":"arwise","name":"ArWISE Vol4","tagline":"CASAS free-living smartwatch activity labels","summary":"ArWISE (CASAS / Diane Cook) is a multi-volume smartwatch free-living activity dataset. Sensor streams are 10 Hz accelerometer, gyroscope, and relative location (distance from home); portions carry user activity labels. Volume 4 DOI 10.5281/zenodo.15809156.","signals":["accel","gyro","hr"],"subjects":46,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/arwise","links":[{"href":"https://zenodo.org/records/15809156","label":"Download","ariaLabel":"Download for ArWISE Vol4 (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.15809156","label":"DOI","ariaLabel":"DOI for ArWISE Vol4 (opens in new tab)"}]},{"slug":"assemblyhar-wrist-acc-manual-assembly","name":"AssemblyHAR","tagline":null,"summary":"AssemblyHAR contains wrist-worn triaxial accelerometer data from 16 participants across 26 sessions of two manual assembly processes. The release includes raw and preprocessed sensor data, activity labels, BLE connection logs, metadata, XES event logs and ready-made windowed datasets, so it can be used for activity recognition and for process mining.","signals":["accel"],"subjects":null,"license":"CC BY 4.0","gated":null,"updated":"2026-10-14","url":"/research/datasets/assemblyhar-wrist-acc-manual-assembly","links":[{"href":"https://doi.org/10.35097/46ntdksar70pfmmn","label":"DOI","ariaLabel":"DOI for AssemblyHAR (opens in new tab)"}]},{"slug":"auricle","name":"Auricle","tagline":"multimodal cardiovascular set: ECG, finger + ear PPG, SCG, in-ear audio, PCG, CNAP + arm cuff (70 subj incl. 28 heart failure, 1 kHz)","summary":"Benchmark built to compare wearable cardiovascular monitors on one common, high-rate dataset. All sensors run on one BIOPAC MP160 at 10 kHz, then downsampled to 1 kHz. Stressors change hemodynamics: breath-hold ×4, light exercise ×2, cold pressor ×1 (healthy only), Valsalva ×4, in a Latin-square order.","signals":["ppg","ecg","accel","audio","abp","resp"],"subjects":70,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/auricle","links":[{"href":"https://physionet.org/content/auricle/1.0.0/","label":"Download","ariaLabel":"Download for Auricle (opens in new tab)"},{"href":"https://doi.org/10.13026/swme-vp69","label":"DOI","ariaLabel":"DOI for Auricle (opens in new tab)"}]},{"slug":"bandx-activity","name":"bandx-activity","tagline":null,"summary":"BandX-Activity is an activity recognition dataset from 33 volunteers wearing the BandX wristband, which uses an MPU-6050 sensor to record accelerometer and gyroscope data at 20 Hz. It includes each person's age, gender, height and weight, so models can be personalised or evaluated by demographic group. A processed version uses 2-second windows with 50% overlap, and sample notebooks show preprocessing and a 1D CNN baseline.","signals":["accel","gyro"],"subjects":33,"license":"CC BY 4.0","gated":null,"updated":"2026-10-14","url":"/research/datasets/bandx-activity","links":[{"href":"https://doi.org/10.5281/zenodo.14543551","label":"DOI","ariaLabel":"DOI for bandx-activity (opens in new tab)"}]},{"slug":"barocontrol-healthy","name":"Barocontrol in healthy subjects (Mol/van Wezel, Radboud)","tagline":"wrist+finger PPG, ECG, NIRS + Finapres (n=34); subset of Radboud PTT","summary":"Collection Barocontrol in healthy subjects (version 1, published 2020-02-24), Donders Institute / Radboud University, di.dcn.DSC6200245101149. Authors: Arjen Mol, Andrea Maier, Richard J.A. van Wezel, Carel Meskers.","signals":["ppg","ecg","abp"],"subjects":34,"license":"Not stated","gated":null,"updated":"2026-10-16","url":"/research/datasets/barocontrol-healthy","links":[{"href":"https://data.ru.nl/collections/di/dcn/DSC_62002451_01_149","label":"Download","ariaLabel":"Download for Barocontrol in healthy subjects (Mol/van Wezel, Radboud) (opens in new tab)"},{"href":"https://doi.org/10.34973/7myp-f567","label":"DOI","ariaLabel":"DOI for Barocontrol in healthy subjects (Mol/van Wezel, Radboud) (opens in new tab)"}]},{"slug":"bed-bcg-carlson","name":"Bed-Based BCG (Carlson, K-State)","tagline":"40 adults, bed BCG + ECG + PPG + Finometer PRO continuous BP (IEEE DataPort OA)","summary":"This dataset contains ballistocardiograms from 40 participants recorded with a custom bed-based system of four electromechanical films and four load cells. Each recording is time-aligned with ECG and PPG from a GE Datex CardioCap 5 monitor and continuous blood pressure waveforms from a Finometer PRO. It was collected at Kansas State University.","signals":["bcg","ecg","ppg","resp","abp"],"subjects":40,"license":"CC BY 4.0","gated":null,"updated":"2026-10-14","url":"/research/datasets/bed-bcg-carlson","links":[{"href":"https://ieee-dataport.org/open-access/bed-based-ballistocardiography-dataset","label":"Download","ariaLabel":"Download for Bed-Based BCG (Carlson, K-State) (opens in new tab)"},{"href":"https://doi.org/10.21227/77hc-py84","label":"DOI","ariaLabel":"DOI for Bed-Based BCG (Carlson, K-State) (opens in new tab)"}]},{"slug":"biagetti","name":"biagetti","tagline":null,"summary":"This dataset pairs wrist PPG recorded during motion with accelerometer data from the same wrist, for studying motion artifacts. It was collected at the Polytechnic University of Marche in Ancona, Italy, from 7 subjects. It holds 105 PPG signals (15 per subject) and 105 matching tri-axial accelerometer signals, all sampled at 400 Hz.","signals":["ppg","accel"],"subjects":7,"license":"CC BY 4.0","gated":null,"updated":"2026-10-15","url":"/research/datasets/biagetti","links":[{"href":"https://doi.org/10.1016/j.dib.2019.105044","label":"DOI","ariaLabel":"DOI for biagetti (opens in new tab)"}]},{"slug":"bidmc","name":"BIDMC PPG and Respiration (Pimentel 2016 / PhysioNet)","tagline":null,"summary":"Not re-downloaded.","signals":["ppg","ecg","resp","spo2"],"subjects":53,"license":"ODC-By 1.0","gated":false,"updated":"2026-10-15","url":"/research/datasets/bidmc","links":[{"href":"https://physionet.org/content/bidmc/1.0.0/","label":"Download","ariaLabel":"Download for BIDMC PPG and Respiration (Pimentel 2016 / PhysioNet) (opens in new tab)"}]},{"slug":"bidsleep","name":"BIDSleep","tagline":"Apple Watch IHR + accel with Dreem EEG sleep stages","summary":"BIDSleep (PhysioNet bidsleep-dataset) pairs consumer smartwatch instantaneous heart rate (IHR) and wrist accelerometry with expert-reviewed Dreem 2 EEG hypnograms. Authors: Song / Zhang / Zhou / Dutta (UMass Lowell). Companion paper: IEEE TBME 2025 (10.1109/TBME.2025.3612158).","signals":["accel","hr","eeg"],"subjects":47,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/bidsleep","links":[{"href":"https://physionet.org/content/bidsleep-dataset/1.0.1/","label":"Download","ariaLabel":"Download for BIDSleep (opens in new tab)"},{"href":"https://doi.org/10.13026/rees-1092","label":"DOI","ariaLabel":"DOI for BIDSleep (opens in new tab)"}]},{"slug":"big-ideas-glycemic","name":"BIG IDEAs Lab Glycemic Variability and Wearable Device Data","tagline":null,"summary":"PhysioNet big-ideas-glycemic-wearable v1.1.3 (DOI 10.13026/aw6y-fc44, published April 13, 2026). Continuous free-living wrist PPG with interstitial glucose.","signals":["ppg","accel","skin-temp","eda","cgm"],"subjects":16,"license":"Open Data Commons Attribution License v1.0","gated":false,"updated":"2026-10-14","url":"/research/datasets/big-ideas-glycemic","links":[{"href":"https://physionet.org/content/big-ideas-glycemic-wearable/1.1.3/","label":"Download","ariaLabel":"Download for BIG IDEAs Lab Glycemic Variability and Wearable Device Data (opens in new tab)"},{"href":"https://doi.org/10.13026/aw6y-fc44","label":"DOI","ariaLabel":"DOI for BIG IDEAs Lab Glycemic Variability and Wearable Device Data (opens in new tab)"}]},{"slug":"biostress","name":"BIOSTRESS (psychologist induction + exam)","tagline":null,"summary":"Psychological-stress dataset pairing a psychologist-led induction session with later exam recordings (Data-in-Brief 2023). Mendeley DOI 10.17632/z9y2fn38fp.1. Therapy folders prefixed T, exam folders prefixed E.","signals":["ppg","accel","skin-temp","eda"],"subjects":20,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/biostress","links":[{"href":"https://data.mendeley.com/datasets/z9y2fn38fp/1","label":"Download","ariaLabel":"Download for BIOSTRESS (psychologist induction + exam) (opens in new tab)"},{"href":"https://doi.org/10.17632/z9y2fn38fp.1","label":"DOI","ariaLabel":"DOI for BIOSTRESS (psychologist induction + exam) (opens in new tab)"}]},{"slug":"blue-green-ppg-shchelkanova","name":"Blue and Green PPG (Shchelkanova / Mendeley)","tagline":null,"summary":"Shchelkanova et al., Kyushu Institute of Technology (Jun 2021). Includes derived heartbeat intervals + LabVIEW software.","signals":["ppg","ecg"],"subjects":12,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/blue-green-ppg-shchelkanova","links":[{"href":"https://data.mendeley.com/datasets/6m8kmdv8ys/1","label":"Download","ariaLabel":"Download for Blue and Green PPG (Shchelkanova / Mendeley) (opens in new tab)"},{"href":"https://doi.org/10.17632/6m8kmdv8ys.1","label":"DOI","ariaLabel":"DOI for Blue and Green PPG (Shchelkanova / Mendeley) (opens in new tab)"}]},{"slug":"blumio-cv-sensors","name":"Blumio","tagline":"115 adults, finger PPG + radial tonometry + wrist mmWave radar vs CNAP continuous BP (IEEE DataPort OA)","summary":"The Blumio dataset was collected to support work on continuous, non-invasive blood pressure monitoring with wearable sensors. It includes PPG with blood pressure reference data recorded under a set protocol with subjects at rest.","signals":["ppg","abp"],"subjects":115,"license":"CC BY 4.0","gated":null,"updated":"2026-10-14","url":"/research/datasets/blumio-cv-sensors","links":[{"href":"https://ieee-dataport.org/open-access/dataset-synchronized-signals-wearable-cardiovascular-monitoring-sensors","label":"Download","ariaLabel":"Download for Blumio (opens in new tab)"},{"href":"https://doi.org/10.21227/3yte-wz05","label":"DOI","ariaLabel":"DOI for Blumio (opens in new tab)"}]},{"slug":"bmd-hs","name":"BMD-HS","tagline":"BUET multi-disease heart sound dataset","summary":"BMD-HS is a heart sound dataset for automated diagnosis of valve disease, recorded at the National Institute of Cardiovascular Disease in Dhaka, Bangladesh. It covers aortic stenosis, aortic regurgitation, mitral regurgitation, mitral stenosis, multiple disease and normal, with multi-label annotations, severity and de-identified demographics. The training folder holds 872 recordings from 59 patients at 8 chest positions, each 20 seconds long at 4 kHz.","signals":["audio"],"subjects":null,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/bmd-hs","links":[{"href":"https://github.com/sani002/BMD-HS-Dataset","label":"Download","ariaLabel":"Download for BMD-HS (opens in new tab)"},{"href":"https://doi.org/10.1016/j.cmpbup.2026.100237","label":"DOI","ariaLabel":"DOI for BMD-HS (opens in new tab)"}]},{"slug":"brist1d","name":"BrisT1D-Open","tagline":"longitudinal T1D smartwatch study","summary":"BrisT1D was conducted by Sam Gordon James, Miranda Armstrong, Aisling O'Kane, Harry Emerson, and Zahraa Abdallah at the University of Bristol. Participants used smartwatches alongside their T1D self-management and donated smartwatch/T1D-device data during the six-month study. The BrisT1D-Open record contains the processed state; the raw state is a separate restricted-access dataset.","signals":["hr","cgm","steps","distance","calories","activity-derived"],"subjects":24,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/brist1d","links":[{"href":"https://data.bris.ac.uk/data/dataset/33z5jc8fa6tob21ptrugzqog08","label":"Download","ariaLabel":"Download for BrisT1D-Open (opens in new tab)"},{"href":"https://doi.org/10.5523/bris.33z5jc8fa6tob21ptrugzqog08","label":"DOI","ariaLabel":"DOI for BrisT1D-Open (opens in new tab)"}]},{"slug":"butppg","name":"Brno University of Technology Smartphone PPG Database (BUT PPG)","tagline":null,"summary":null,"signals":["ppg","ecg"],"subjects":50,"license":"CC BY 4.0","gated":null,"updated":"2026-10-14","url":"/research/datasets/butppg","links":[{"href":"https://physionet.org/content/butppg/2.0.0/","label":"Download","ariaLabel":"Download for Brno University of Technology Smartphone PPG Database (BUT PPG) (opens in new tab)"},{"href":"https://doi.org/10.13026/tn53-8153","label":"DOI","ariaLabel":"DOI for Brno University of Technology Smartphone PPG Database (BUT PPG) (opens in new tab)"}]},{"slug":"cachet-cadb","name":"CACHET-CADB: A Contextualized Ambulatory Electrocardiography Arrhythmia Dataset","tagline":null,"summary":"24 patients with known or suspected paroxysmal AF: 7 Danish and 17 Indian, 15 M / 9 F, mean age 59. They wore a Movisens EcgMove4 chest Holter (mCardia system, REAFEL study) in free living for 24 h to 3 weeks, giving 259 days in total. Raw signals: ECG at 1024 Hz, ACC and gyroscope at 64 Hz, pressure at 8 Hz.","signals":["ecg","accel","gyro","barometer"],"subjects":null,"license":"CC BY 4.0","gated":null,"updated":"2026-10-14","url":"/research/datasets/cachet-cadb","links":[{"href":"https://doi.org/10.11583/DTU.14547264","label":"DOI","ariaLabel":"DOI for CACHET-CADB: A Contextualized Ambulatory Electrocardiography Arrhythmia Dataset (opens in new tab)"}]},{"slug":"can-stress","name":"CAN-STRESS (ambulatory E4 stress 0–9)","tagline":null,"summary":"Real-world ambulatory multimodal stress + cannabis-use dataset. Zenodo DOI 10.5281/zenodo.14842061. ~24 h wear during daily activities.","signals":["ppg","accel","skin-temp","eda"],"subjects":82,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/can-stress","links":[{"href":"https://zenodo.org/records/14842061","label":"Download","ariaLabel":"Download for CAN-STRESS (ambulatory E4 stress 0–9) (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.14842061","label":"DOI","ariaLabel":"DOI for CAN-STRESS (ambulatory E4 stress 0–9) (opens in new tab)"}]},{"slug":"capnobase","name":"CapnoBase IEEE TBME RR benchmark (Karlen 2013)","tagline":null,"summary":"The CapnoBase TBME RR benchmark contains 42 eight-minute recordings with PPG and respiration, used to compare respiratory rate algorithms on the same data. The authors ask that it not be used to train or tune algorithms, because that would bias the comparison.","signals":["ppg","ecg","resp"],"subjects":null,"license":"Custom terms specific to this dataset","gated":null,"updated":"2026-10-16","url":"/research/datasets/capnobase","links":[{"href":"https://doi.org/10.5683/SP2/NLB8IT","label":"Download","ariaLabel":"Download for CapnoBase IEEE TBME RR benchmark (Karlen 2013) (opens in new tab)"}]},{"slug":"capture-24","name":"capture-24","tagline":null,"summary":"CAPTURE-24 (OxWearables / Oxford): free-living wrist accelerometry with wearable-camera activity annotations. Nature Scientific Data descriptor; ORA release.","signals":["accel"],"subjects":151,"license":"CC BY 4.0","gated":null,"updated":"2026-10-14","url":"/research/datasets/capture-24","links":[{"href":"https://ora.ox.ac.uk/objects/uuid:99d7c092-d865-4a19-b096-cc16440cd001","label":"Download","ariaLabel":"Download for capture-24 (opens in new tab)"},{"href":"https://doi.org/10.5287/bodleian:NGx0JOMP5","label":"DOI","ariaLabel":"DOI for capture-24 (opens in new tab)"}]},{"slug":"cardiowave-portable","name":"CardioWave-Portable","tagline":"ECG+PPG cuffless BP (n=979, 3857×24s)","summary":"CardioWave-Portable (Yang / Zhang / Liu / Wang / Xie / Guo; Harbin Institute of Technology affiliation on Zenodo). Zenodo record 20912312 (concept 20799109), published 2026-06-26. Release zip includes waveforms, data dictionary, loading code, and P1/P2/P3 benchmark scripts.","signals":["ppg","ecg"],"subjects":979,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/cardiowave-portable","links":[{"href":"https://zenodo.org/records/20912312","label":"Download","ariaLabel":"Download for CardioWave-Portable (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.20912312","label":"DOI","ariaLabel":"DOI for CardioWave-Portable (opens in new tab)"}]},{"slug":"chapman-shaoxing-ningbo","name":"Chapman–Shaoxing–Ningbo 12-lead ECG arrhythmia database (PhysioNet)","tagline":null,"summary":"10-s resting 12-lead ECGs from 45,152 patients (Chapman University, Shaoxing People's Hospital, Ningbo First Hospital), GE MUSE; rhythm labels (incl. AFIB) and conditions (PVC, RBBB, LBBB, APB) by two physicians, SNOMED-CT coded. Earlier 10,646-patient subset on figshare (10.6084/m9.figshare.c.4560497.v2).","signals":["ecg"],"subjects":45152,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/chapman-shaoxing-ningbo","links":[{"href":"https://physionet.org/content/ecg-arrhythmia/1.0.0/","label":"Download","ariaLabel":"Download for Chapman–Shaoxing–Ningbo 12-lead ECG arrhythmia database (PhysioNet) (opens in new tab)"},{"href":"https://doi.org/10.13026/wgex-er52","label":"DOI","ariaLabel":"DOI for Chapman–Shaoxing–Ningbo 12-lead ECG arrhythmia database (PhysioNet) (opens in new tab)"}]},{"slug":"challenge-2015","name":"CinC / PhysioNet Challenge 2015","tagline":"Reducing false arrhythmia alarms in the ICU","summary":"Official 2015 PhysioNet/CinC Challenge corpus (Clifford et al.). Training set: 750 recordings with expert true/false alarm labels; hidden test 500 remains private. Each record leads up to an alarm at t≈5 min; retrospective subset adds 30 s post-alarm.","signals":["ppg","ecg","abp"],"subjects":null,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/challenge-2015","links":[{"href":"https://physionet.org/content/challenge-2015/1.0.0/","label":"Download","ariaLabel":"Download for CinC / PhysioNet Challenge 2015 (opens in new tab)"}]},{"slug":"clas","name":"CLAS","tagline":"Cognitive Load, Affect and Stress (Shimmer ECG/PPG/EDA)","summary":"CLAS (Cognitive Load, Affect and Stress) supports automated recognition of emotion, stress, and cognitive load. Authors: Markova / Ganchev / Kalinkov (TU Varna). Paper: BIA-2019 10.1109/BIA48344.2019.8967457.","signals":["ppg","ecg","accel","eda"],"subjects":62,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/clas","links":[{"href":"https://ieee-dataport.org/open-access/database-cognitive-load-affect-and-stress-recognition","label":"Download","ariaLabel":"Download for CLAS (opens in new tab)"},{"href":"https://doi.org/10.21227/ybsw-yr53","label":"DOI","ariaLabel":"DOI for CLAS (opens in new tab)"}]},{"slug":"clbp-400","name":"CLBP-400","tagline":"rPPG face video cuffless BP, indoor/outdoor (n=400)","summary":"400 HD face videos at 60 fps (iPhone 16 Pro Max, Nikon D5300, Elgato 4K60), 30-60 s each, ~83 GB, indoor and outdoor with lux values; ground truth SBP/DBP/HR (BP device not stated). Preprint preprints.org 202604.1079. Access by DUA, non-commercial academic only → gated / eval-only.","signals":["ppg"],"subjects":400,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/clbp-400","links":[{"href":"https://github.com/alialnaji/CLBP-300-Dataset","label":"Download","ariaLabel":"Download for CLBP-400 (opens in new tab)"}]},{"slug":"cleveland-family-study","name":"Cleveland Family Study (CFS)","tagline":"NSRR family sleep-apnea PSG","summary":"The Cleveland Family Study is the largest family-based study of sleep apnea, with 2,284 people (46% African American) from 361 families. Participants were studied up to four times between 1990 and 2006, and the final visit included full overnight polysomnography. The data are shared by the National Sleep Research Resource for non-commercial use only.","signals":["psg","ppg"],"subjects":2284,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/cleveland-family-study","links":[{"href":"https://sleepdata.org/datasets/cfs","label":"Download","ariaLabel":"Download for Cleveland Family Study (CFS) (opens in new tab)"},{"href":"https://doi.org/10.25822/jmyx-mz90","label":"DOI","ariaLabel":"DOI for Cleveland Family Study (CFS) (opens in new tab)"}]},{"slug":"code-15","name":"CODE-15%: large annotated 12-lead ECG dataset (Telehealth Network of Minas Gerais)","tagline":null,"summary":"CODE-15% is a 12-lead ECG dataset drawn by stratified sampling of 15% of the patients in the larger CODE dataset. The exams were collected by the Telehealth Network of Minas Gerais in Brazil between 2010 and 2016. Each exam comes with patient age, sex and a neural-network-predicted ECG age, as used in a study of ECG age as a mortality predictor.","signals":["ecg"],"subjects":null,"license":"CC BY 4.0","gated":null,"updated":"2026-10-16","url":"/research/datasets/code-15","links":[{"href":"https://zenodo.org/records/4916206","label":"Download","ariaLabel":"Download for CODE-15%: large annotated 12-lead ECG dataset (Telehealth Network of Minas Gerais) (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.4916206","label":"DOI","ariaLabel":"DOI for CODE-15%: large annotated 12-lead ECG dataset (Telehealth Network of Minas Gerais) (opens in new tab)"}]},{"slug":"feng-imu-layouts-2026","name":"Comprehensive IMU dataset for sensor layouts in HAR + intensity (Feng)","tagline":null,"summary":"Figshare v2 (2025-09-25, Mingfei Feng). Page text gives title only. Awesome list: 30 subjects.","signals":["accel"],"subjects":null,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/feng-imu-layouts-2026","links":[{"href":"https://figshare.com/articles/dataset/Comprehensive-IMU-Dataset/30193942","label":"Download","ariaLabel":"Download for Comprehensive IMU dataset for sensor layouts in HAR + intensity (Feng) (opens in new tab)"}]},{"slug":"bp-graphene-bioimpedance","name":"Continuous cuffless BP via graphene bioimpedance tattoos (PhysioNet)","tagline":null,"summary":"This PhysioNet dataset measures blood pressure without a cuff using atomically thin graphene electronic tattoos that record bioimpedance over the radial and ulnar arteries at the wrist. It includes raw 4-channel bioimpedance with synchronized PPG and reference blood pressure from a Finapres NOVA. Seven people were monitored for about 4 hours each, giving 18,667 beat-level blood pressure data points.","signals":["ppg","abp","bioimpedance"],"subjects":8,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/bp-graphene-bioimpedance","links":[{"href":"https://physionet.org/content/bp-graphene-bioimpedance/1.0.0/","label":"Download","ariaLabel":"Download for Continuous cuffless BP via graphene bioimpedance tattoos (PhysioNet) (opens in new tab)"}]},{"slug":"covid-lockdown-gear-sport","name":"COVID Lockdown Gear Sport (longitudinal wearable + EMA)","tagline":null,"summary":"Intensive longitudinal physiological + emotional assessment of college students before/during/after the 2020 COVID-19 lockdown (Southern California). Dryad DOI 10.7280/D1WH6T. ~8.6 GB.","signals":["ppg","accel"],"subjects":null,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/covid-lockdown-gear-sport","links":[{"href":"https://datadryad.org/dataset/doi:10.7280/D1WH6T","label":"Download","ariaLabel":"Download for COVID Lockdown Gear Sport (longitudinal wearable + EMA) (opens in new tab)"}]},{"slug":"crhe-long-term-ppg-bp","name":"CRHE","tagline":"29/30 elderly, 3-λ (green/red/IR) wearable PPG segments paired with dual-24h ambulatory BP (Fudan, MIT)","summary":"Fudan University (Yuhui Cen, Jingjing Luo) with Univ. of Shanghai for Science and Technology (Hongliu Yu) and Changzhou Anxin Yihe and Rehabilitation Hospital (Yufeng Bao, Bingzhong Huang). Abstract: \"self-collected multi-wavelength, dual-24h PPG-BP dataset … circadian rhythm annotations … collected using wearable devices and ambulatory BP monitors.\" Paper: tpf-bp-physiology-guided-ssl-personalized-cuffless-bp-crhe-2026.","signals":["ppg"],"subjects":29,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/crhe-long-term-ppg-bp","links":[{"href":"https://github.com/FDU-Bio-Monitoring/CRHE-Long-Term-PPG-BP-Dataset","label":"Download","ariaLabel":"Download for CRHE (opens in new tab)"},{"href":"https://doi.org/10.21227/bgnp-xs19","label":"DOI","ariaLabel":"DOI for CRHE (opens in new tab)"}]},{"slug":"d1namo","name":"D1NAMO","tagline":"multimodal type 1 diabetes management","summary":"D1NAMO was collected by the D1NAMO project with authors from the University of Applied Sciences and Arts Western Switzerland, EPFL, Open University of the Netherlands, and Hôpital Riviera-Chablais. The Zenodo record publishes separate healthy and diabetes subsets for ECG, sensor data, and food/glucose annotations. The companion article is documented at 10.1016/j.imu.2018.09.003.","signals":["ecg","accel","resp","cgm"],"subjects":29,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/d1namo","links":[{"href":"https://zenodo.org/records/5651217","label":"Download","ariaLabel":"Download for D1NAMO (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.5651217","label":"DOI","ariaLabel":"DOI for D1NAMO (opens in new tab)"}]},{"slug":"daily-and-sports-activities","name":"daily-and-sports-activities","tagline":null,"summary":"This UCI dataset records 19 daily and sports activities, from sitting and climbing stairs to treadmill running and rowing, each done by 8 people (4 women, 4 men, aged 20 to 30) for 5 minutes in their own style. Five Xsens MTx units on the torso, arms and legs recorded at 25 Hz, and the signals are split into 5-second segments. Data were collected at Bilkent University and donated in 2013.","signals":["accel"],"subjects":8,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/daily-and-sports-activities","links":[{"href":"https://doi.org/10.24432/C5C59F","label":"DOI","ariaLabel":"DOI for daily-and-sports-activities (opens in new tab)"}]},{"slug":"daphnet-fog","name":"Daphnet Freezing of Gait","tagline":"Parkinson's patients, hip/leg accelerometers (n=10)","summary":"Three acceleration sensors at hip and leg of Parkinson's disease patients with freezing of gait; recorded in the lab (Tel Aviv Sourasky Medical Center, 2008): straight walking, turns, and an ADL task. UCI, CC BY 4.0. N=10 and 8.3 h per the original dataset documentation listed on Awesome-IMU-Sensing (not re-checked).","signals":["accel"],"subjects":10,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/daphnet-fog","links":[{"href":"https://archive.ics.uci.edu/dataset/245/daphnet+freezing+of+gait","label":"Download","ariaLabel":"Download for Daphnet Freezing of Gait (opens in new tab)"}]},{"slug":"deo","name":"DEO","tagline":"Drinking / Eating / Other smartwatch IMU (Samsung GW4 relabel)","summary":"Defined in Riveros et al., “Diffusion-Based Synthetic Data Pretraining for Enhancing Activity Recognition” (arXiv 2610.02292, 2026-10-01). Authors start from a Daily Living Activities Dataset (90 participants, 13 activities, Samsung Galaxy Watch 4 wrist IMU) and relabel into DEO: drinking + eating from the dominant hand; other = all non-alimentary activities from the opposite hand.","signals":["accel","gyro","magnetometer"],"subjects":90,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/deo","links":[{"href":"https://arxiv.org/abs/2610.02292","label":"Download","ariaLabel":"Download for DEO (opens in new tab)"}]},{"slug":"dose-i","name":"DOSE-I","tagline":"endoscopy propofol sedation multimodal (EEG+ECG+PLETH+NIBP, n=171 records)","summary":"DOSE-I (Garbe / Kantelhardt / Seeliger / Schmid; Universitätsmedizin Halle). Prospective single-center endoscopy sedation study (DRKS00016605; May 2019–Jul 2021). Export via MIB/RS232 + ixTrend Express.","signals":["ppg","ecg","eeg","resp","spo2","hr"],"subjects":171,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/dose-i","links":[{"href":"https://zenodo.org/records/22145442","label":"Download","ariaLabel":"Download for DOSE-I (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.22145442","label":"DOI","ariaLabel":"DOI for DOSE-I (opens in new tab)"}]},{"slug":"dreamt","name":"DREAMT","tagline":"Dataset for Real-time sleep stage EstimAtion using Multisensor wearable Technology","summary":"Wrist PPG with PSG gold-standard sleep staging — the largest single wrist-PPG cohort in the ledger and the only one with expert-scored overnight sleep labels. Collected at Duke University Health System Sleep Disorder Lab, May–Sept 2022.","signals":["ppg","accel","skin-temp","eda","psg"],"subjects":100,"license":"PhysioNet Restricted Health Data License 1.5.0","gated":true,"updated":"2026-10-16","url":"/research/datasets/dreamt","links":[{"href":"https://physionet.org/content/dreamt/","label":"Download","ariaLabel":"Download for DREAMT (opens in new tab)"}]},{"slug":"e4-stress-exercise","name":"E4 Stress + Exercise (PhysioNet wearable-device-dataset)","tagline":null,"summary":"Empatica E4 wristband recordings under three distinct physiological load conditions: acute psychological stress, aerobic exercise, anaerobic exercise. Published June 2025, PhysioNet DOI 10.13026/he0v-tf17, companion paper Scientific Data 12(1) 10.1038/s41597-025-04845-9. Member of e4-commercial-cluster.","signals":["ppg","accel","skin-temp","eda"],"subjects":36,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/e4-stress-exercise","links":[{"href":"https://physionet.org/content/wearable-device-dataset/1.0.1/","label":"Download","ariaLabel":"Download for E4 Stress + Exercise (PhysioNet wearable-device-dataset) (opens in new tab)"},{"href":"https://doi.org/10.13026/he0v-tf17","label":"DOI","ariaLabel":"DOI for E4 Stress + Exercise (PhysioNet wearable-device-dataset) (opens in new tab)"}]},{"slug":"earset","name":"EarSet","tagline":"multi-modal in-ear green/red/IR PPG + IMU","summary":"EarSet is a multimodal dataset for studying how body, head and face movement affects in-ear PPG wave shape and vital-sign estimates. The authors built an in-ear research platform that records PPG at several wavelengths from both ears together with 6-axis motion, 18 data streams in total. Recordings cover facial and head movements as well as walking and running.","signals":["ppg","ecg","accel","gyro","resp"],"subjects":30,"license":"CC BY 4.0","gated":null,"updated":"2026-10-15","url":"/research/datasets/earset","links":[{"href":"https://zenodo.org/records/8142332","label":"Download","ariaLabel":"Download for EarSet (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.8142332","label":"DOI","ariaLabel":"DOI for EarSet (opens in new tab)"}]},{"slug":"ecsmp","name":"ECSMP","tagline":"Emotion, Cognition, Sleep Multimodal Physiological Signals","summary":"ECSMP contains EEG, ECG, PPG, electrodermal activity, temperature and accelerometer recordings from 89 healthy college students. They were recorded at rest, during emotion induction and recovery, and during cognitive tests. Sleep ECG was taken the night before, and questionnaires covered emotion regulation, depression, mood and sleep quality.","signals":["ppg","ecg","accel","skin-temp","eda","eeg"],"subjects":89,"license":"CC BY 4.0","gated":null,"updated":"2026-10-14","url":"/research/datasets/ecsmp","links":[{"href":"https://data.mendeley.com/datasets/vn5nknh3mn/2","label":"Download","ariaLabel":"Download for ECSMP (opens in new tab)"},{"href":"https://doi.org/10.17632/vn5nknh3mn.2","label":"DOI","ariaLabel":"DOI for ECSMP (opens in new tab)"}]},{"slug":"eevr","name":"EEVR","tagline":"VR emotion elicitation with PPG + EDA and textual emotion descriptions (NeurIPS 2024 D&B)","summary":"Pragya Singh, Ritvik Budhiraja, Ankush Gupta, Anshul Goswami, Mohan Kumar, Pushpendra Singh (IIIT-Delhi; RIT). NeurIPS 2024 Datasets & Benchmarks (OpenReview qgzdGyQcDt). Introduces CLSP (contrastive language–signal pretraining).","signals":["ppg","eda"],"subjects":37,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/eevr","links":[{"href":"https://melangelabiiitd.github.io/EEVR/","label":"Download","ariaLabel":"Download for EEVR (opens in new tab)"}]},{"slug":"emowear","name":"EmoWear (E4 SAM affect)","tagline":null,"summary":"Wearable physiological + motion dataset for emotion recognition / context awareness (Scientific Data 2024). Affect-adjacent to stress (SAM valence/arousal/dominance), not a dedicated stress protocol. Zenodo DOI 10.5281/zenodo.10407279.","signals":["ppg","accel","skin-temp","eda"],"subjects":48,"license":"Not stated","gated":null,"updated":"2026-10-16","url":"/research/datasets/emowear","links":[{"href":"https://zenodo.org/records/10407279","label":"Download","ariaLabel":"Download for EmoWear (E4 SAM affect) (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.10407279","label":"DOI","ariaLabel":"DOI for EmoWear (E4 SAM affect) (opens in new tab)"}]},{"slug":"e4-commercial-cluster","name":"Empatica E4 Commercially-Clear Cluster (CogWear, Exam Stress, ScientISST MOVE, Neuro-Stress-Resilience-HCI, Wrist PPG During Exercise)","tagline":null,"summary":"Five separate PhysioNet datasets that individually are too small to matter but share an almost identical sensor layout, so they concatenate into one corpus with a single extraction pipeline.","signals":["ppg","accel","skin-temp","eda","ecg-ref"],"subjects":null,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/e4-commercial-cluster","links":[{"href":"https://physionet.org/about/database/","label":"Download","ariaLabel":"Download for Empatica E4 Commercially-Clear Cluster (CogWear, Exam Stress, ScientISST MOVE, Neuro-Stress-Resilience-HCI, Wrist PPG During Exercise) (opens in new tab)"}]},{"slug":"empatica-e4-stress","name":"EmpaticaE4Stress (workplace cognitive stress)","tagline":null,"summary":"Workplace-style cognitive / social / psychological stress protocol with Empatica E4 (Campanella et al.). Mendeley Data v2, DOI 10.17632/kb42z77m2g.2. Companion Data-in-Brief 2024.","signals":["ppg","accel","skin-temp","eda"],"subjects":29,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/empatica-e4-stress","links":[{"href":"https://data.mendeley.com/datasets/kb42z77m2g/2","label":"Download","ariaLabel":"Download for EmpaticaE4Stress (workplace cognitive stress) (opens in new tab)"},{"href":"https://doi.org/10.17632/kb42z77m2g.2","label":"DOI","ariaLabel":"DOI for EmpaticaE4Stress (workplace cognitive stress) (opens in new tab)"}]},{"slug":"emteq-breathing-vr","name":"Emteq Breathing Data VR (forehead PPG)","tagline":null,"summary":"Kaggle emteqlabs/breathingdatavr. Three sub-datasets (free + controlled breathing rates). Paper: Stankoski et al., Sensors 2022.","signals":["ppg","accel"],"subjects":40,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/emteq-breathing-vr","links":[{"href":"https://www.kaggle.com/datasets/emteqlabs/breathingdatavr","label":"Download","ariaLabel":"Download for Emteq Breathing Data VR (forehead PPG) (opens in new tab)"},{"href":"https://doi.org/10.3390/s22062079","label":"DOI","ariaLabel":"DOI for Emteq Breathing Data VR (forehead PPG) (opens in new tab)"}]},{"slug":"epistress","name":"EPIStress (epilepsy cognitive stress, E4+Muse)","tagline":null,"summary":"EPIStress records EEG, PPG, acceleration, electrodermal activity and temperature while people with epilepsy did cognitively stressful and relaxing tasks. It is meant for studying the link between stress and seizures that patients often report.","signals":["ppg","accel","skin-temp","eda","eeg"],"subjects":20,"license":"CC BY 4.0","gated":null,"updated":"2026-10-14","url":"/research/datasets/epistress","links":[{"href":"https://zenodo.org/records/16407549","label":"Download","ariaLabel":"Download for EPIStress (epilepsy cognitive stress, E4+Muse) (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.16407549","label":"DOI","ariaLabel":"DOI for EPIStress (epilepsy cognitive stress, E4+Muse) (opens in new tab)"}]},{"slug":"eppg-glucose-stu-zenodo","name":"ePPG glucose dataset (STU Bratislava)","tagline":"6-channel 3-wavelength (530/655/940 nm) thumb PPG + ACC with fingerstick glucose (n=24, CC0)","summary":"Berki & Mateašík (Slovak University of Technology / International Laser Center). Zenodo 10.5281/zenodo.21308261 (2026-07-11; later versions 10.5281/zenodo.21978226 and 22206256 exist).","signals":["ppg","accel"],"subjects":24,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/eppg-glucose-stu-zenodo","links":[{"href":"https://doi.org/10.5281/zenodo.21308261","label":"Download","ariaLabel":"Download for ePPG glucose dataset (STU Bratislava) (opens in new tab)"}]},{"slug":"escolar","name":"ESCOLAR","tagline":"EmotiBit academic stress + cognitive load","summary":"ESCOLAR integrates two domains on EmotiBit: social stress in naturalistic academic presentations and controlled programming cognitive-load tasks. Zenodo 19495742, DOI 10.5281/zenodo.19495742, license CC BY 4.0. >1.2M synchronized instances claimed; includes processing scripts.","signals":["ppg","accel","gyro","skin-temp","eda"],"subjects":26,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/escolar","links":[{"href":"https://zenodo.org/records/19495742","label":"Download","ariaLabel":"Download for ESCOLAR (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.19495742","label":"DOI","ariaLabel":"DOI for ESCOLAR (opens in new tab)"}]},{"slug":"eviora-har","name":"EVIORA-HAR","tagline":"Elderly multi-view + 5-node wearable IMU for ADL and safety-controlled falls","summary":"EVIORA-HAR (Elderly Multi-View and Wearable Motion Recognition Dataset) from Iwan Kurnianto Wibowo, Chastine Fatichah, and Nanik Suciati (Institut Teknologi Sepuluh Nopember / PENS). Zenodo mirror https://zenodo.org/records/22239523 carries IEEE DataPort DOI 10.21227/2zvc-nb24 (publication date on record 2026-08-09; Zenodo created 2026-09-02). Companion Data in Brief article DOI 10.1016/j.dib.2026.113285 (PMC13654039; PubMed 42860996).","signals":["accel","gyro"],"subjects":104,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/eviora-har","links":[{"href":"https://zenodo.org/records/22239523","label":"Download","ariaLabel":"Download for EVIORA-HAR (opens in new tab)"},{"href":"https://doi.org/10.21227/2zvc-nb24","label":"DOI","ariaLabel":"DOI for EVIORA-HAR (opens in new tab)"}]},{"slug":"falltl","name":"FallTL","tagline":"8-site IMU pre-/post-impact fall dataset (28.4 h)","summary":"FallTL is a wearable inertial dataset for detecting falls both before and after impact. Sensors were placed at eight body locations and recorded 12 everyday activities and 34 types of falls, about 28.43 hours in total, with ready-made fall labels. It was published on Zenodo in November 2025 by authors from Dalian University of Technology and partner institutions.","signals":["accel"],"subjects":null,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/falltl","links":[{"href":"https://zenodo.org/records/17552449","label":"Download","ariaLabel":"Download for FallTL (opens in new tab)"}]},{"slug":"finger-contact-ecg-dual-wavelength-ppg-post-exercise","name":"Finger-contact ECG + dual-wavelength (660/940 nm) PPG, rest + post-exercise recovery (n=148, 444 records)","tagline":null,"summary":"This dataset has synchronized single-lead ECG and dual-wavelength PPG (660 and 940 nm) from 148 healthy subjects, recorded with a handheld finger-contact prototype. Each subject gave three seated records: one at rest, one after light treadmill running and one after moderate running, for 444 records in total.","signals":["ppg","ecg"],"subjects":148,"license":"CC BY 4.0","gated":null,"updated":"2026-10-14","url":"/research/datasets/finger-contact-ecg-dual-wavelength-ppg-post-exercise","links":[{"href":"https://doi.org/10.3390/data11070155","label":"Download","ariaLabel":"Download for Finger-contact ECG + dual-wavelength (660/940 nm) PPG, rest + post-exercise recovery (n=148, 444 records) (opens in new tab)"}]},{"slug":"galaxy-active2-mental-health","name":"Galaxy Watch Active2 raw PPG+ACC (PHQ-9 / GAD-7)","tagline":null,"summary":"This dataset contains four weeks of continuous smartwatch recordings from 49 healthy people (mean age 28.4, 51% women), sampled every 100 ms so heart rate variability can be computed for each 5-minute segment. Participants also kept daily sleep diaries and completed clinical questionnaires every two weeks on anxiety, depression and insomnia. It is meant for benchmarking real-world HRV and predicting mental health and sleep.","signals":["ppg","accel"],"subjects":null,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/galaxy-active2-mental-health","links":[{"href":"https://doi.org/10.6084/m9.figshare.28509740","label":"Download","ariaLabel":"Download for Galaxy Watch Active2 raw PPG+ACC (PHQ-9 / GAD-7) (opens in new tab)"}]},{"slug":"galaxyppg","name":"GalaxyPPG","tagline":null,"summary":"Semi-naturalistic wrist PPG dataset collected at KAIST with a consumer smartwatch and a research-grade wristband recording simultaneously, plus a chest ECG reference. Published as a Scientific Data descriptor (2025), DOI 10.1038/s41597-025-05152-z.","signals":["ppg","ecg","accel","skin-temp"],"subjects":24,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/galaxyppg","links":[{"href":"https://doi.org/10.5281/zenodo.14635823","label":"Download","ariaLabel":"Download for GalaxyPPG (opens in new tab)"}]},{"slug":"groupaffect4","name":"GroupAffect-4","tagline":"EmotiBit multimodal four-person collaboration","summary":"GroupAffect-4 is a multimodal corpus of co-located four-person collaboration (T0 rest → T1 hidden-profile → T2 negotiation → T3 ideation → T4 public-goods). Creators include Jamshidi Seikavandi et al. Full release ~30 GB (Zenodo 20037847); subset release without raw audio (Zenodo 20809799, recommended start).","signals":["ppg","accel","skin-temp","eda","audio","eye-tracking"],"subjects":40,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/groupaffect4","links":[{"href":"https://zenodo.org/records/20809799","label":"Download","ariaLabel":"Download for GroupAffect-4 (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.20809799","label":"DOI","ariaLabel":"DOI for GroupAffect-4 (opens in new tab)"}]},{"slug":"gyro-acc-ppg-bami1","name":"Gyro-Acc-PPG (BAMI-I, Wonkwang)","tagline":"3-channel green wrist PPG + accel + gyro, treadmill","summary":"Hooseok Lee, Heewon Chung, Jinseok Lee (BAMI Lab, Wonkwang University). Paper: \"Motion Artifact Cancellation in Wearable Photoplethysmography Using Gyroscope\", IEEE Sensors Journal, doi 10.1109/JSEN.2018.2879970. GitHub repo created 2018-10-05; 24 Subject.mat files.","signals":["ppg","accel","gyro","ecg"],"subjects":24,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/gyro-acc-ppg-bami1","links":[{"href":"https://github.com/hooseok/gyro_acc_ppg","label":"Download","ariaLabel":"Download for Gyro-Acc-PPG (BAMI-I, Wonkwang) (opens in new tab)"},{"href":"https://doi.org/10.1109/JSEN.2018.2879970","label":"DOI","ariaLabel":"DOI for Gyro-Acc-PPG (BAMI-I, Wonkwang) (opens in new tab)"}]},{"slug":"har70plus","name":"HAR70+","tagline":"older adults (70-95 y), thigh + back accelerometers (n=18)","summary":"HAR70+ is a professionally annotated activity dataset from 18 older adults aged 70 to 95, ranging from fit to frail, five of whom used walking aids. Each wore two Axivity AX3 three-axis accelerometers, on the right front thigh and the lower back, for about 40 minutes of semi-structured free-living activity. Data are provided at 50 Hz and were donated to the UCI repository in 2023.","signals":["accel"],"subjects":18,"license":"Creative Commons Attribution 4.0 International (CC BY 4.0)","gated":false,"updated":"2026-10-14","url":"/research/datasets/har70plus","links":[{"href":"https://archive.ics.uci.edu/dataset/780/har70","label":"Download","ariaLabel":"Download for HAR70+ (opens in new tab)"}]},{"slug":"harmes","name":"HARMES","tagline":"wrist IMU + atmospheric + audio HAR","summary":"HARMES (Burchard / Brückner / Bock / Van Laerhoven et al.) — multi-modal wearable HAR with motion, environmental sensing, and sound. Zenodo 10.5281/zenodo.23416820 (concept 19425718; prior record 19425719); arXiv 2605.02596; code https://github.com/RBurchard/HARMES. Raw + 50 Hz preprocessed pickles.","signals":["accel","gyro","audio","atmospheric-temp","humidity","pressure"],"subjects":20,"license":"Not stated","gated":null,"updated":"2026-10-16","url":"/research/datasets/harmes","links":[{"href":"https://zenodo.org/records/23416820","label":"Download","ariaLabel":"Download for HARMES (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.23416820","label":"DOI","ariaLabel":"DOI for HARMES (opens in new tab)"}]},{"slug":"harth","name":"HARTH","tagline":"thigh + lower-back accelerometers, free-living, video-annotated (n=22)","summary":"HARTH (Human Activity Recognition Trondheim) contains professionally annotated free-living recordings from 22 participants. Each wore two Axivity AX3 three-axis accelerometers, on the right front thigh and lower back, for about two hours. It was funded by NTNU Helse and donated to the UCI repository in 2023 as a benchmark for free-living activity recognition.","signals":["accel"],"subjects":22,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/harth","links":[{"href":"https://archive.ics.uci.edu/dataset/779/harth","label":"Download","ariaLabel":"Download for HARTH (opens in new tab)"}]},{"slug":"healthring","name":"HealthRing RingDatasetV2.1","tagline":"finger-ring IR/red PPG + IMU with HR/RR/SpO2/cuff-BP refs (τ-Ring successor)","summary":"Tang, Wang, Ding, Ji, Wang, Wang, Zhang, Chen, Gao, Shi, Wang. HealthRing: Physiology Dataset for Health Sensing on Rings. Sci Data (2026), doi:10.1038/s41597-026-07289-x.","signals":["ppg","accel","ecg","resp"],"subjects":54,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/healthring","links":[{"href":"https://zenodo.org/records/18426864","label":"Download","ariaLabel":"Download for HealthRing RingDatasetV2.1 (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.18426864","label":"DOI","ariaLabel":"DOI for HealthRing RingDatasetV2.1 (opens in new tab)"}]},{"slug":"heart-watch","name":"HEART-Watch","tagline":"Pixel Watch wrist ECG+PPG+accel, chest-ECG reference, intermittent cuff BP (n=40)","summary":"HEART-Watch is a dataset of synchronized ECG, PPG and accelerometer signals from a Google Pixel Watch, with reference chest ECG. It covers 40 healthy adults from a deliberately diverse cohort, recorded while sitting, standing and walking. The authors made it because public smartwatch datasets with synchronized cardiac signals are scarce and often lack demographic diversity.","signals":["ppg","ecg","accel"],"subjects":40,"license":"CC BY 4.0","gated":null,"updated":"2026-10-16","url":"/research/datasets/heart-watch","links":[{"href":"https://doi.org/10.1371/journal.pdig.0001647","label":"Download","ariaLabel":"Download for HEART-Watch (opens in new tab)"}]},{"slug":"heartcycle","name":"HeartCycle: A comprehensive dataset of synchronized impedance cardiography and echocardiography for accurate hemodynamic predictions","tagline":null,"summary":"HeartCycle pairs impedance cardiography with echocardiography, plus finger PPG, heart sounds and ECG, from 17 healthy volunteers. It was collected to study biases in the ICG waveform, where the B and X points do not line up exactly with aortic valve opening and closing. Those points are used to compute systolic time intervals, contractility, stroke volume and cardiac output.","signals":["ppg","ecg"],"subjects":null,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/heartcycle","links":[{"href":"https://physionet.org/content/heartcycle/1.0.0/","label":"Download","ariaLabel":"Download for HeartCycle: A comprehensive dataset of synchronized impedance cardiography and echocardiography for accurate hemodynamic predictions (opens in new tab)"},{"href":"https://doi.org/10.13026/z865-eb23","label":"DOI","ariaLabel":"DOI for HeartCycle: A comprehensive dataset of synchronized impedance cardiography and echocardiography for accurate hemodynamic predictions (opens in new tab)"}]},{"slug":"hhar","name":"HHAR","tagline":"heterogeneity HAR, smartphones + smartwatches (n=9)","summary":"The Heterogeneity Human Activity Recognition (HHAR) dataset was built to benchmark activity recognition when the same activities are recorded on many different smartphone and smartwatch models. It contains accelerometer and gyroscope readings captured while users biked, sat, stood, walked and climbed stairs, and it is meant for testing classification, segmentation, sensor fusion and feature extraction under real device differences. It was donated to the UCI repository in 2015 and has about 43.9 million instances.","signals":["accel"],"subjects":9,"license":"Creative Commons Attribution 4.0 International (CC BY 4.0)","gated":false,"updated":"2026-10-14","url":"/research/datasets/hhar","links":[{"href":"https://archive.ics.uci.edu/dataset/344/heterogeneity+activity+recognition","label":"Download","ariaLabel":"Download for HHAR (opens in new tab)"}]},{"slug":"human-sleep-project","name":"Human Sleep Project (BDSP)","tagline":"multi-site clinical PSG","summary":"Restricted BDSP dataset; access needs BDSP credentialing plus the Credentialed Health DUA. v3.0 published 2026-06-02 (Li, Wen, Sun, Ganglberger, … Westover), DOI 10.60508/m3sw-rz13; earlier v2.0 DOI 10.60508/qjbv-hg78 (cited by AnyPPG). Data in BIDS layout with HDF5 signals, linked to de-identified demographics and EHR (22 ICD-10 categories).","signals":["psg","ppg"],"subjects":90000,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/human-sleep-project","links":[{"href":"https://bdsp.io/content/hsp/3.0/","label":"Download","ariaLabel":"Download for Human Sleep Project (BDSP) (opens in new tab)"},{"href":"https://doi.org/10.60508/m3sw-rz13","label":"DOI","ariaLabel":"DOI for Human Sleep Project (BDSP) (opens in new tab)"}]},{"slug":"icbhi-2017","name":"ICBHI 2017 Respiratory Sound Database","tagline":null,"summary":"The ICBHI 2017 database is a public set of respiratory sound recordings created for a scientific challenge. It contains 920 recordings from 126 participants with two sets of annotations, one for respiratory cycles and one for adventitious events. The best challenge entry scored 52.5% on the cycle annotations and 91.2% on the event annotations.","signals":["audio"],"subjects":126,"license":"Not stated","gated":false,"updated":"2026-10-15","url":"/research/datasets/icbhi-2017","links":[{"href":"https://bhichallenge.med.auth.gr/ICBHI_2017_Challenge","label":"Download","ariaLabel":"Download for ICBHI 2017 Respiratory Sound Database (opens in new tab)"},{"href":"https://doi.org/10.1088/1361-6579/ab03ea","label":"DOI","ariaLabel":"DOI for ICBHI 2017 Respiratory Sound Database (opens in new tab)"}]},{"slug":"icentia11k","name":"Icentia11k single-lead continuous raw ECG (PhysioNet)","tagline":null,"summary":"Icentia11k is a large set of continuous raw single-lead ECG from 11,000 patients, with 2 billion labelled beats. Signals were recorded at 250 Hz with 16-bit resolution by a chest-mounted single-lead device worn for up to two weeks.","signals":["ecg"],"subjects":null,"license":"Not stated","gated":null,"updated":"2026-10-16","url":"/research/datasets/icentia11k","links":[{"href":"https://physionet.org/content/icentia11k-continuous-ecg/1.0/","label":"Download","ariaLabel":"Download for Icentia11k single-lead continuous raw ECG (PhysioNet) (opens in new tab)"},{"href":"https://doi.org/10.13026/kk0v-r952","label":"DOI","ariaLabel":"DOI for Icentia11k single-lead continuous raw ECG (PhysioNet) (opens in new tab)"}]},{"slug":"ieee-spc-2015-troika","name":"IEEE Signal Processing Cup 2015 (TROIKA)","tagline":"wrist PPG during intensive exercise","summary":"This is the IEEE Signal Processing Cup 2015 heart rate dataset, also hosted in the Monash/UEA/UCR time series regression repository. Two wrist PPG channels (green LEDs, 515 nm), three-axis wrist acceleration and one ECG channel were recorded together from subjects aged 18 to 35. The task is to estimate heart rate from PPG during movement.","signals":["ppg","accel","ecg"],"subjects":12,"license":"CC BY 4.0","gated":null,"updated":"2026-10-15","url":"/research/datasets/ieee-spc-2015-troika","links":[{"href":"https://signalprocessingsociety.org/community-involvement/ieee-signal-processing-cup-2015","label":"Download","ariaLabel":"Download for IEEE Signal Processing Cup 2015 (TROIKA) (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.3902710","label":"DOI","ariaLabel":"DOI for IEEE Signal Processing Cup 2015 (TROIKA) (opens in new tab)"}]},{"slug":"imu50","name":"IMU50","tagline":"50-subject 5-day ActiGraph Leap wrist PPG+IMU","summary":"IMU50 (La Fratta / Calisti / Bigelli et al.): multiday high-resolution wrist IMU+PPG under unrestricted free-living. Zenodo concept DOI 10.5281/zenodo.21468409 (version DOI 10.5281/zenodo.21468410). Intended for (1) self-supervised IMU representation learning and (2) supervised metabolic / HR / morphology tasks.","signals":["ppg","accel","gyro","skin-temp"],"subjects":50,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/imu50","links":[{"href":"https://zenodo.org/records/21468409","label":"Download","ariaLabel":"Download for IMU50 (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.21468410","label":"DOI","ariaLabel":"DOI for IMU50 (opens in new tab)"}]},{"slug":"infrared-bp-tomstog","name":"Infrared-BP","tagline":"IR camera forehead + palm video, PTT-based BP (small n)","summary":"Infrared-BP is a dataset of infrared-camera videos of the forehead and upper palm, used to estimate blood pressure without contact. The pipeline applies Eulerian video magnification, estimates pulse transit time and fits the result to reference blood pressure values. It supports a CNNA 2023 paper and a 2024 MDPI BioMedInformatics paper, and the videos are downloadable from MEGA.","signals":["ppg"],"subjects":null,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/infrared-bp-tomstog","links":[{"href":"https://github.com/TomStog/Infrared-BP","label":"Download","ariaLabel":"Download for Infrared-BP (opens in new tab)"}]},{"slug":"iomt-secalarmbench","name":"IoMT-SecAlarmBench (semi-synthetic ECG+PPG attack benchmark)","tagline":null,"summary":"IoMT-SecAlarmBench is a benchmark for detecting integrity attacks on medical device data. Controlled attacks are injected into real paired ECG and PPG recordings from the CinC 2015 ICU alarm challenge, and the clean signal is kept for every attacked window. It has 78,729 labelled 10-second windows at 125 Hz with patient-wise splits, and no tested detector reliably caught replay attacks or small transient spikes.","signals":["ppg","ecg"],"subjects":null,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/iomt-secalarmbench","links":[{"href":"https://github.com/techosystem/IoMT_SecAlarmBench","label":"Download","ariaLabel":"Download for IoMT-SecAlarmBench (semi-synthetic ECG+PPG attack benchmark) (opens in new tab)"}]},{"slug":"ipqm-fall","name":"IPqM-Fall","tagline":"Galaxy Watch 4 + phone IMU fall/ADL/military","summary":"IPqM-Fall is a fall dataset recorded from 15 military personnel at the Navy Research Institute during daily and military activities, including weapon use. Data came from a Samsung Galaxy Watch 4 and an LG Velvet smartphone and include linear acceleration on three axes.","signals":["accel","gyro"],"subjects":15,"license":"CC BY 4.0","gated":null,"updated":"2026-10-14","url":"/research/datasets/ipqm-fall","links":[{"href":"https://zenodo.org/records/12760391","label":"Download","ariaLabel":"Download for IPqM-Fall (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.12760391","label":"DOI","ariaLabel":"DOI for IPqM-Fall (opens in new tab)"}]},{"slug":"iridia-af","name":"IRIDIA-AF","tagline":"long-term Holter ECG with expert AF annotations","summary":"Gilon et al., Scientific Data companion release (Nature Sci Data article linked from Zenodo). Zenodo record 8405941 v1.0.1 (accessright=open, license.id=cc-by-4.0). GitHub: cedricgilon/iridia-af.","signals":["ecg"],"subjects":152,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/iridia-af","links":[{"href":"https://zenodo.org/records/8405941","label":"Download","ariaLabel":"Download for IRIDIA-AF (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.8405941","label":"DOI","ariaLabel":"DOI for IRIDIA-AF (opens in new tab)"}]},{"slug":"ispl-imu-stretch","name":"iSPL IMU-Stretch","tagline":"single subject, wrist/waist/ankle IMU + knee stretch","summary":"The iSPL IMU-Stretch dataset comes from one person wearing three IMUs (right wrist, waist and ankle) and two fabric stretch sensors at the knees. It covers nine activities, including walking, running, push-ups and dancing, and the data are pre-windowed with a 70/30 train/test split. It accompanies an ICAIIC 2022 paper on deep learning activity recognition with IMUs and stretch sensors.","signals":["accel"],"subjects":1,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/ispl-imu-stretch","links":[{"href":"https://github.com/thunguyenth/HAR_IMU_Stretch","label":"Download","ariaLabel":"Download for iSPL IMU-Stretch (opens in new tab)"}]},{"slug":"k-emocon","name":"K-EmoCon","tagline":"multimodal emotion in naturalistic debates (n=32, Empatica E4)","summary":"Park et al., Scientific Data 7:293 (2020). Zenodo record v1.0.0 DOI 10.5281/zenodo.3931963 (files restricted; user agreement https://forms.gle/DUAERhHqf51kyt4Y9). KAIST IRB.","signals":["ppg","accel","eda","skin-temp","ecg","eeg","hr"],"subjects":32,"license":"Restricted access (request on Zenodo); no license listed","gated":true,"updated":"2026-10-15","url":"/research/datasets/k-emocon","links":[{"href":"https://zenodo.org/records/3931963","label":"Download","ariaLabel":"Download for K-EmoCon (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.3931963","label":"DOI","ariaLabel":"DOI for K-EmoCon (opens in new tab)"}]},{"slug":"kingston-icu-af","name":"Kingston ICU AF","tagline":"clinician-labelled 10 s 4-lead ICU ECGs","summary":"Nassar / Maghsoodi / Chen / Maslove / Sibley / Laird / Mousavi (Queen’s / KHSC). Archival ICU monitors 2015–2020; one random 10 s strip per labelled patient. Companion TBME 2026 benchmarks: 10.1109/TBME.2026.3715145.","signals":["ecg"],"subjects":596,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/kingston-icu-af","links":[{"href":"https://physionet.org/content/kingston-icu-af/1.0.0/","label":"Download","ariaLabel":"Download for Kingston ICU AF (opens in new tab)"},{"href":"https://doi.org/10.13026/47wc-qk15","label":"DOI","ariaLabel":"DOI for Kingston ICU AF (opens in new tab)"}]},{"slug":"landry-adl-bp","name":"Landry ADL","tagline":"5 healthy adults × 6.5–8.4 h free-living, vest PPG/ECG + Portapres continuous finger BP (IEEE DataPort)","summary":"University of Waterloo (Landry, Hedge, Hughson, Peterson, Arami). Collected to validate a cuffless BP model during ADL. Paper: \"Accurate Blood Pressure Estimation during Activities of Daily Living: A Wearable Cuffless Solution\", IEEE J-BHI 2021.","signals":["ppg","ecg","accel","abp"],"subjects":5,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/landry-adl-bp","links":[{"href":"https://ieee-dataport.org/documents/wearable-physiological-and-blood-pressure-measurements-during-activities-daily-living","label":"Download","ariaLabel":"Download for Landry ADL (opens in new tab)"},{"href":"https://doi.org/10.21227/wysp-gt69","label":"DOI","ariaLabel":"DOI for Landry ADL (opens in new tab)"}]},{"slug":"lara-logistics","name":"LARa v03","tagline":"logistics HAR, IMU + OMoCap + RGB (n=16)","summary":"TU Dortmund logistics lab: two picking and one packing scenario, 16 subjects, optical marker MoCap, IMUs and RGB camera; each subject recorded 1 h (960 min total); 8 activity classes + 19 attributes. Zenodo, CC BY-NC 4.0, 12 files, 42.7 GB.","signals":["accel"],"subjects":16,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/lara-logistics","links":[{"href":"https://zenodo.org/records/8189341","label":"Download","ariaLabel":"Download for LARa v03 (opens in new tab)"}]},{"slug":"lightweight-bp-xt2j","name":"Lightweight BP Estimation (Yongjian Li)","tagline":"300 subjects N/P/S1, 1 s PPG/ECG/ABP windows at 125 Hz, provenance unknown (IEEE DataPort)","summary":"This dataset supports a lightweight blood pressure estimation study using PPG and ECG. Each group contains 100 subjects with records of at least 8 minutes, which the study segments into 1-second windows with a 1-second step.","signals":["ppg","ecg","abp"],"subjects":300,"license":"CC BY 4.0","gated":null,"updated":"2026-10-16","url":"/research/datasets/lightweight-bp-xt2j","links":[{"href":"https://ieee-dataport.org/documents/lightweight-blood-pressure-estimation-algorithm","label":"Download","ariaLabel":"Download for Lightweight BP Estimation (Yongjian Li) (opens in new tab)"},{"href":"https://doi.org/10.21227/xt2j-5t27","label":"DOI","ariaLabel":"DOI for Lightweight BP Estimation (Yongjian Li) (opens in new tab)"}]},{"slug":"loneliness-fi","name":"Loneliness-FI","tagline":"Finnish immigrants wearable multimodal (Dryad)","summary":"Loneliness-FI was collected from first-generation immigrants between 2022 and 2023. Data came from a Samsung Watch Active 2, an Oura Ring, the AWARE phone sensing app and the Centralive smartphone app.","signals":["ppg","accel","gyro"],"subjects":39,"license":"CC0 1.0","gated":null,"updated":"2026-10-15","url":"/research/datasets/loneliness-fi","links":[{"href":"https://datadryad.org/dataset/doi:10.5061/dryad.qz612jmrn","label":"Download","ariaLabel":"Download for Loneliness-FI (opens in new tab)"},{"href":"https://doi.org/10.5061/dryad.qz612jmrn","label":"DOI","ariaLabel":"DOI for Loneliness-FI (opens in new tab)"}]},{"slug":"ltafdb","name":"Long-Term AF Database (LTAFDB, PhysioNet)","tagline":null,"summary":"The Long-Term AF Database contains 84 long-term ECG recordings from people with paroxysmal or sustained atrial fibrillation. Each record has two ECG signals at 128 Hz and 12-bit resolution, and most last 24 to 25 hours.","signals":["ecg"],"subjects":null,"license":"Not stated","gated":null,"updated":"2026-10-16","url":"/research/datasets/ltafdb","links":[{"href":"https://physionet.org/content/ltafdb/1.0.0/","label":"Download","ariaLabel":"Download for Long-Term AF Database (LTAFDB, PhysioNet) (opens in new tab)"},{"href":"https://doi.org/10.13026/C2QG6Q","label":"DOI","ariaLabel":"DOI for Long-Term AF Database (LTAFDB, PhysioNet) (opens in new tab)"}]},{"slug":"long-term-af-ppg-ecg-bacevicius-2024","name":"Long-term ECG + wrist PPG with annotated AF episodes (Bacevičius 2024, Zenodo)","tagline":null,"summary":"This dataset contains long-term wrist PPG recorded together with reference ECG, with atrial fibrillation episodes annotated. Because there are no guidelines for reading arrhythmias from PPG, the ECG is there to confirm episodes, which makes the data suited to building and testing PPG-based AF detectors.","signals":["ppg","ecg","accel"],"subjects":45,"license":"Other (Non-Commercial)","gated":null,"updated":"2026-10-16","url":"/research/datasets/long-term-af-ppg-ecg-bacevicius-2024","links":[{"href":"https://zenodo.org/records/23396020","label":"Download","ariaLabel":"Download for Long-term ECG + wrist PPG with annotated AF episodes (Bacevičius 2024, Zenodo) (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.23396020","label":"DOI","ariaLabel":"DOI for Long-term ECG + wrist PPG with annotated AF episodes (Bacevičius 2024, Zenodo) (opens in new tab)"}]},{"slug":"afppgecg","name":"Long-term ECG and wrist PPG with annotated AF episodes (AFPPGECG)","tagline":null,"summary":"45 patients with suspected atrial fibrillation, monitored 5–8 days each (306 monitoring days total), starting at Vilnius University Hospital and continuing at home. Wrist PPG device running continuously alongside a clinical ECG patch. Zenodo record v2, June 2024 (DOI 10.5281/zenodo.11242869).","signals":["ppg","ecg","accel"],"subjects":39,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/afppgecg","links":[{"href":"https://zenodo.org/records/23187681","label":"Download","ariaLabel":"Download for Long-term ECG and wrist PPG with annotated AF episodes (AFPPGECG) (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.23187681","label":"DOI","ariaLabel":"DOI for Long-term ECG and wrist PPG with annotated AF episodes (AFPPGECG) (opens in new tab)"}]},{"slug":"lown2020-polar-af","name":"Lown 2020 Polar H7","tagline":"RR-interval AF screening (PLOS ONE)","summary":"This PLOS ONE study looked at whether a cheap consumer ECG chest strap, the Polar H7, can be used to detect atrial fibrillation. Such devices measure ECG but usually lack diagnostic algorithms.","signals":["rr-interval"],"subjects":415,"license":"CC BY 4.0","gated":null,"updated":"2026-10-15","url":"/research/datasets/lown2020-polar-af","links":[{"href":"https://doi.org/10.1371/journal.pone.0227401","label":"Download","ariaLabel":"Download for Lown 2020 Polar H7 (opens in new tab)"}]},{"slug":"maibo","name":"MAIBO","tagline":"bilateral ring PPG multi-posture cuff BP (n≈1086 released / 1810 advertised)","summary":"MAIBO (Chen / Li / Liu / Wang / Shi / Gao / Qiu / Tu / Wang / Wang / Dai / Chen; Shanghai Jiao Tong University + Mianyang Central Hospital). Zenodo record 19440421 (published 2026-04-06, V1). Companion Sci Data descriptor: A Multi-posture Asymmetry-aware Intelligent Bilateral Observation Dataset for Cardiovascular Monitoring (doi:10.1038/s41597-026-07915-8).","signals":["ppg"],"subjects":1086,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/maibo","links":[{"href":"https://zenodo.org/records/19440421","label":"Download","ariaLabel":"Download for MAIBO (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.19440421","label":"DOI","ariaLabel":"DOI for MAIBO (opens in new tab)"}]},{"slug":"maus-mental-workload","name":"MAUS","tagline":"Mental Workload Assessment (N-back) with Wearable Sensors","summary":"MAUS (Mental workloAd assessment Under n-back task using wearable Sensors) is a multimodal physiological dataset for mental workload (MW) assessment. Subjects perform an N-back working-memory task while clinical-grade and consumer wrist sensors record in parallel. Authors: Beh / Wu / Wu (NTU).","signals":["ecg","eda","ibi","finger-ppg","wrist-ppg"],"subjects":22,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/maus-mental-workload","links":[{"href":"https://ieee-dataport.org/open-access/maus-dataset-mental-workload-assessment-n-back-task-using-wearable-sensor","label":"Download","ariaLabel":"Download for MAUS (opens in new tab)"},{"href":"https://doi.org/10.21227/q4td-yd35","label":"DOI","ariaLabel":"DOI for MAUS (opens in new tab)"}]},{"slug":"mayo-lbnp-hypovolemia","name":"Mayo LBNP hypovolemia: non-invasive biomarkers during lower body negative pressure","tagline":null,"summary":"This is the raw data behind a study of non-invasive biomarkers for detecting progression toward hypovolemic cardiovascular instability in a lower body negative pressure model. It covers 16 subjects with ECG, mean arterial pressure and pleth (PPG) recordings, plus electrical impedance and bioimpedance cardiography data.","signals":["ppg","ecg","bp","bioimpedance"],"subjects":16,"license":"CC BY 4.0","gated":null,"updated":"2026-10-15","url":"/research/datasets/mayo-lbnp-hypovolemia","links":[{"href":"https://doi.org/10.5281/zenodo.10119428","label":"Download","ariaLabel":"Download for Mayo LBNP hypovolemia: non-invasive biomarkers during lower body negative pressure (opens in new tab)"}]},{"slug":"mbb-avr-affective-vr-multimodal","name":"MBB-AVR","tagline":"MindBrainBody affective VR (EEG, ECG, PPG, GSR, resp; 47 subj)","summary":"MBB-AVR records continuous emotion ratings alongside brain and body signals while people watched immersive VR. Forty-seven healthy adults viewed a 23-minute 360-degree video on an HTC Vive Pro Eye headset and rated valence and arousal in real time. Recordings include 64-channel EEG, ECG, respiration, PPG, skin conductance, and head and eye tracking with pupillometry.","signals":["eeg","ecg","ppg","eda","resp","eye-tracking"],"subjects":47,"license":"CC BY 4.0","gated":null,"updated":"2026-10-14","url":"/research/datasets/mbb-avr-affective-vr-multimodal","links":[{"href":"https://doi.org/10.31234/osf.io/vg2qc_v1","label":"Download","ariaLabel":"Download for MBB-AVR (opens in new tab)"}]},{"slug":"mcphases-menstrual-wearable","name":"mcPHASES: A Dataset of Physiological, Hormonal, and Self-reported Events and Symptoms for Menstrual Health Tracking with Wearables","tagline":null,"summary":"42 Canadian young adults, two 3-month intervals, Fitbit Sense + Dexcom G6 CGM + Mira Plus urinary hormone analyzer + daily symptom diary. Published via PhysioNet 2025-09-09, University of Toronto REB-approved.","signals":["skin-temp","spo2","hrv","fitbit-derived heart_rate","sleep_stages","respiratory_rate","activity","vo2max","stress_score","continuous_glucose","urinary_hormones"],"subjects":null,"license":"Not stated","gated":null,"updated":"2026-10-16","url":"/research/datasets/mcphases-menstrual-wearable","links":[{"href":"https://physionet.org/content/mcphases/","label":"Download","ariaLabel":"Download for mcPHASES: A Dataset of Physiological, Hormonal, and Self-reported Events and Symptoms for Menstrual Health Tracking with Wearables (opens in new tab)"}]},{"slug":"mefar","name":"MEFAR","tagline":"Occupational Mental Fatigue (E4 BVP+ACC + Chalder)","summary":"Workplace / occupational mental fatigue multimodal set released with Data in Brief (2023/2024). Morning and evening sessions on the same workers; Chalder Fatigue Scale labels paired with Empatica E4 waveforms and single-channel EEG. Mendeley DOI 10.17632/z3g26tphnv.5; DiB DOI 10.1016/j.dib.2023.109896.","signals":["ppg","accel","skin-temp","eda","eeg"],"subjects":23,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/mefar","links":[{"href":"https://data.mendeley.com/datasets/z3g26tphnv/5","label":"Download","ariaLabel":"Download for MEFAR (opens in new tab)"},{"href":"https://doi.org/10.17632/z3g26tphnv.5","label":"DOI","ariaLabel":"DOI for MEFAR (opens in new tab)"}]},{"slug":"mesa","name":"MESA Sleep","tagline":"NSRR polysomnography + wrist actigraphy (Multi-Ethnic Study of Atherosclerosis)","summary":"Multi-Ethnic Study of Atherosclerosis: 6,814 Black, White, Hispanic and Chinese-American adults aged 45–84 at baseline. The sleep ancillary study is distributed by NSRR with PSG and actigraphy raw data plus covariates.","signals":["psg","ppg","ecg","eeg","resp","accel"],"subjects":2237,"license":"Not stated","gated":null,"updated":"2026-10-16","url":"/research/datasets/mesa","links":[{"href":"https://sleepdata.org/datasets/mesa","label":"Download","ariaLabel":"Download for MESA Sleep (opens in new tab)"},{"href":"https://doi.org/10.25822/n7hq-c406","label":"DOI","ariaLabel":"DOI for MESA Sleep (opens in new tab)"}]},{"slug":"mfd-multimodal-driver-fatigue","name":"MFD","tagline":"Multimodal Driver Fatigue: EEG, ECG, EMG, EOG, PPG, BCG + tri-view RGB-D face, PERCLOS labels (19 subj x 2 sessions x 2 h)","summary":"MFD is a driving fatigue dataset with six synchronized physiological signals (EEG, ECG, EMG, EOG, PPG and BCG) and three-view RGB-D video of the face. Nineteen healthy participants each drove for two hours in a fixed scenario across two sessions. Fatigue levels were labelled with PERCLOS, an objective eyelid-closure measure.","signals":["eeg","ecg","eog","ppg","bcg"],"subjects":19,"license":"CC BY 4.0","gated":null,"updated":"2026-10-14","url":"/research/datasets/mfd-multimodal-driver-fatigue","links":[{"href":"https://ieee-dataport.org/documents/mfd-multimodal-driver-fatigue","label":"Download","ariaLabel":"Download for MFD (opens in new tab)"},{"href":"https://doi.org/10.21227/25yw-mc95","label":"DOI","ariaLabel":"DOI for MFD (opens in new tab)"}]},{"slug":"mgmcet-fmci-pbf","name":"MGMCET-FMCI-PBF","tagline":"IPG/PPG/ECG peripheral blood flow DB (India); Study-4 held: 73 × 5 min ECG + index-finger PPG, 19 hypertensive","summary":"This dataset contains impedance plethysmography (IPG) and PPG recordings used to estimate peripheral blood flow. Study 1 has 14,478 five-minute wrist IPG records from about 200 young healthy volunteers, recorded before and after a placebo or homeopathic medicines at several potencies.","signals":["ecg","ppg"],"subjects":73,"license":"CC BY 4.0","gated":null,"updated":"2026-10-14","url":"/research/datasets/mgmcet-fmci-pbf","links":[{"href":"https://ieee-dataport.org/documents/mgmcet-fmci-pbf-peripheral-blood-flow-database-using-ipg-ppg-and-ecg","label":"Download","ariaLabel":"Download for MGMCET-FMCI-PBF (opens in new tab)"},{"href":"https://doi.org/10.21227/rbwb-zx09","label":"DOI","ariaLabel":"DOI for MGMCET-FMCI-PBF (opens in new tab)"}]},{"slug":"mhealth","name":"mhealth","tagline":null,"summary":"MHEALTH records body motion and vital signs from ten volunteers doing 12 physical activities. Shimmer2 sensors on the chest, right wrist and left ankle measured acceleration, rate of turn and magnetic field, and the chest sensor also recorded 2-lead ECG. It was donated to the UCI repository in 2014 to benchmark human behavior analysis from body-worn sensors.","signals":["accel"],"subjects":10,"license":"Creative Commons Attribution 4.0 International (CC BY 4.0)","gated":false,"updated":"2026-10-14","url":"/research/datasets/mhealth","links":[{"href":"https://archive.ics.uci.edu/dataset/319/mhealth+dataset","label":"Download","ariaLabel":"Download for mhealth (opens in new tab)"},{"href":"https://doi.org/10.24432/C5TW22","label":"DOI","ariaLabel":"DOI for mhealth (opens in new tab)"}]},{"slug":"mimic-perform","name":"MIMIC PERform","tagline":"Charlton PPG/ECG/RESP ICU beat-detection benchmarks (+ AF subset)","summary":"Charlton PERform family used in PPG beat-detector benchmarking (Physiol. Meas. 2022, DOI 10.1088/1361-6579/ac826d).","signals":["ppg","ecg","resp","abp"],"subjects":35,"license":"ODbL","gated":false,"updated":"2026-10-15","url":"/research/datasets/mimic-perform","links":[{"href":"https://zenodo.org/records/15906524","label":"Download","ariaLabel":"Download for MIMIC PERform (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.15906524","label":"DOI","ariaLabel":"DOI for MIMIC PERform (opens in new tab)"}]},{"slug":"mimic-bp","name":"MIMIC-BP","tagline":"1,524 ICU subjects × 30 × 30 s ABP/PPG/ECG/RESP (Samsung, ODbL)","summary":"MIMIC-BP is derived from the MIMIC-III Waveform Database Matched Subset. It holds 380 hours of arterial blood pressure, PPG and ECG from 1,524 de-identified patients, with 30 segments of 30 seconds each.","signals":["ppg","abp","ecg","resp"],"subjects":1524,"license":"Open Data Commons Open Database License v1.0.","gated":null,"updated":"2026-10-15","url":"/research/datasets/mimic-bp","links":[{"href":"https://doi.org/10.7910/DVN/DBM1NF","label":"Download","ariaLabel":"Download for MIMIC-BP (opens in new tab)"}]},{"slug":"mimic-matched-bp","name":"MIMIC-III Waveform MATCHED","tagline":"paired PLETH+ABP windows","summary":null,"signals":["ppg","abp"],"subjects":2854,"license":"ODbL 1.0","gated":false,"updated":"2026-10-16","url":"/research/datasets/mimic-matched-bp","links":[{"href":"https://physionet.org/content/mimic3wdb-matched/1.0/","label":"Download","ariaLabel":"Download for MIMIC-III Waveform MATCHED (opens in new tab)"}]},{"slug":"mimic-iii-ext-ca","name":"MIMIC-III-Ext-CA","tagline":"PPG cardiac-arrest episode annotations (n=31, 36 events)","summary":"MIMIC-III-Ext-CA is a PPG dataset for detecting out-of-hospital cardiac arrest, built from MIMIC-III. Public data for this task are scarce, and the dataset is meant for developing and testing wearable cardiac arrest detection.","signals":["ppg","ecg"],"subjects":31,"license":"Not stated","gated":null,"updated":"2026-10-16","url":"/research/datasets/mimic-iii-ext-ca","links":[{"href":"https://physionet.org/content/mimic-iii-ext-ca/1.0.0/","label":"Download","ariaLabel":"Download for MIMIC-III-Ext-CA (opens in new tab)"},{"href":"https://doi.org/10.13026/ec0n-y377","label":"DOI","ariaLabel":"DOI for MIMIC-III-Ext-CA (opens in new tab)"}]},{"slug":"mimic-iii-ext-ppg","name":"MIMIC-III-Ext-PPG","tagline":"large ICU PPG rhythm/BP/SQI benchmark (6.3M×30s)","summary":"Curated extract from MIMIC-III Waveform Database Matched Subset (Moulaeifard / Charlton / Strodthoff). Segments end at rhythm chart events (≤15 min lookback). Sci Data companion: 10.1038/s41597-026-07335-8.","signals":["ppg","ecg","resp","abp"],"subjects":6189,"license":"PhysioNet Credentialed Health Data License 1.5.0","gated":true,"updated":"2026-10-15","url":"/research/datasets/mimic-iii-ext-ppg","links":[{"href":"https://physionet.org/content/mimic-iii-ext-ppg/1.1.0/","label":"Download","ariaLabel":"Download for MIMIC-III-Ext-PPG (opens in new tab)"},{"href":"https://doi.org/10.13026/r6k1-xt76","label":"DOI","ariaLabel":"DOI for MIMIC-III-Ext-PPG (opens in new tab)"}]},{"slug":"mimic-iv-waveform","name":"MIMIC-IV Waveform Database (PhysioNet preview)","tagline":null,"summary":"The MIMIC-IV Waveform Database collects signals and measurements from ICU bedside monitors, including ECG, PPG, respiration and invasive and non-invasive blood pressure. Combined with the MIMIC-IV clinical records, it gives a detailed view of the physiology of critically ill patients at a single hospital. Version 0.1.0 is an open-access technical preview on PhysioNet.","signals":["ppg","ecg","resp","abp"],"subjects":198,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/mimic-iv-waveform","links":[{"href":"https://physionet.org/content/mimic4wdb/0.1.0/","label":"Download","ariaLabel":"Download for MIMIC-IV Waveform Database (PhysioNet preview) (opens in new tab)"}]},{"slug":"mimic-iv-ext-ecg-glucose","name":"MIMIC-IV-Ext-ECG-Glucose: Temporally Aligned 12-Lead ECG and Laboratory Blood Glucose Pairs with Quality and Pharmacological Annotations","tagline":null,"summary":"This dataset pairs laboratory blood glucose results from MIMIC-IV with the nearest 12-lead diagnostic ECG recorded within 240 minutes, to support research on estimating glucose from ECG. Records come from emergency, inpatient and ICU encounters, and 80,598 of them (18.4%) were taken during an ICU stay.","signals":["ecg"],"subjects":null,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/mimic-iv-ext-ecg-glucose","links":[{"href":"https://doi.org/10.13026/nts2-vj76","label":"DOI","ariaLabel":"DOI for MIMIC-IV-Ext-ECG-Glucose: Temporally Aligned 12-Lead ECG and Laboratory Blood Glucose Pairs with Quality and Pharmacological Annotations (opens in new tab)"}]},{"slug":"mimic3-af-bashar","name":"MIMIC3-AF-Bashar","tagline":"physician AF annotations + MIMIC-III matched waveforms","summary":"Annotations originate from Bashar et al. (JBHI / UConn) Figshare 10.6084/m9.figshare.12149091.v1 (public AF labels for MIMIC-III matched sepsis ECG review).","signals":["ppg","ecg","resp","abp"],"subjects":70,"license":"CC BY 4.0","gated":null,"updated":"2026-10-16","url":"/research/datasets/mimic3-af-bashar","links":[{"href":"https://doi.org/10.6084/m9.figshare.12149091.v1","label":"Download","ariaLabel":"Download for MIMIC3-AF-Bashar (opens in new tab)"}]},{"slug":"mission1bp-gmu-kime","name":"MISSION1BP-GMU-KIME","tagline":"16-ch multi-λ PPG + ECG + ACC cuffless BP (n=1000, GATED)","summary":"MISSION1BP–GMU-KIME (Y. Zhang et al.; companion preprint DOI 10.21203/rs.3.rs-9176759/v1). Zenodo record 21974809, published 2026-08-17.","signals":["ppg","ecg","accel"],"subjects":1000,"license":"Not stated","gated":null,"updated":"2026-10-16","url":"/research/datasets/mission1bp-gmu-kime","links":[{"href":"https://zenodo.org/records/21974809","label":"Download","ariaLabel":"Download for MISSION1BP-GMU-KIME (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.21974809","label":"DOI","ariaLabel":"DOI for MISSION1BP-GMU-KIME (opens in new tab)"}]},{"slug":"mitdb","name":"MIT-BIH Arrhythmia Database (PhysioNet)","tagline":null,"summary":"The MIT-BIH Arrhythmia Database contains 48 half-hour, two-channel ambulatory ECG recordings from 47 subjects, collected at Boston's Beth Israel Hospital between 1975 and 1979. Some excerpts were chosen at random and others to include rarer but clinically important arrhythmias. Experts annotated about 110,000 beats, and the database has been used for arrhythmia detection research since 1980.","signals":["ecg"],"subjects":null,"license":"Not stated","gated":null,"updated":"2026-10-16","url":"/research/datasets/mitdb","links":[{"href":"https://physionet.org/content/mitdb/1.0.0/","label":"Download","ariaLabel":"Download for MIT-BIH Arrhythmia Database (PhysioNet) (opens in new tab)"},{"href":"https://doi.org/10.13026/C2F305","label":"DOI","ariaLabel":"DOI for MIT-BIH Arrhythmia Database (PhysioNet) (opens in new tab)"}]},{"slug":"afdb","name":"MIT-BIH Atrial Fibrillation Database (PhysioNet)","tagline":null,"summary":"The MIT-BIH Atrial Fibrillation Database contains 25 long-term ECG recordings from people with atrial fibrillation, mostly paroxysmal. Each recording is 10 hours long with two ECG channels sampled at 250 samples per second and 12-bit resolution, and rhythm annotations mark the AF episodes. It was published on PhysioNet by George Moody and Roger Mark and is a standard benchmark for AF detection.","signals":["ecg"],"subjects":null,"license":"Not stated","gated":null,"updated":"2026-10-16","url":"/research/datasets/afdb","links":[{"href":"https://physionet.org/content/afdb/1.0.0/","label":"Download","ariaLabel":"Download for MIT-BIH Atrial Fibrillation Database (PhysioNet) (opens in new tab)"},{"href":"https://doi.org/10.13026/C2MW2D","label":"DOI","ariaLabel":"DOI for MIT-BIH Atrial Fibrillation Database (PhysioNet) (opens in new tab)"}]},{"slug":"nsrdb","name":"MIT-BIH Normal Sinus Rhythm Database (PhysioNet)","tagline":null,"summary":"The MIT-BIH Normal Sinus Rhythm Database contains 18 long-term ECG recordings of people referred to the arrhythmia lab at Boston's Beth Israel Hospital who turned out to have no significant arrhythmias. The subjects were 5 men aged 26 to 45 and 13 women aged 20 to 50.","signals":["ecg"],"subjects":null,"license":"Not stated","gated":null,"updated":"2026-10-16","url":"/research/datasets/nsrdb","links":[{"href":"https://physionet.org/content/nsrdb/1.0.0/","label":"Download","ariaLabel":"Download for MIT-BIH Normal Sinus Rhythm Database (PhysioNet) (opens in new tab)"},{"href":"https://doi.org/10.13026/C2NK5R","label":"DOI","ariaLabel":"DOI for MIT-BIH Normal Sinus Rhythm Database (PhysioNet) (opens in new tab)"}]},{"slug":"svdb","name":"MIT-BIH Supraventricular Arrhythmia Database (PhysioNet)","tagline":null,"summary":"The MIT-BIH Supraventricular Arrhythmia Database contains 78 half-hour ECG recordings. They were chosen to add more examples of supraventricular arrhythmias to the MIT-BIH Arrhythmia Database, and the set is open access on PhysioNet under the Open Data Commons Attribution License.","signals":["ecg"],"subjects":null,"license":"Not stated","gated":null,"updated":"2026-10-16","url":"/research/datasets/svdb","links":[{"href":"https://physionet.org/content/svdb/1.0.0/","label":"Download","ariaLabel":"Download for MIT-BIH Supraventricular Arrhythmia Database (PhysioNet) (opens in new tab)"},{"href":"https://doi.org/10.13026/C2V30W","label":"DOI","ariaLabel":"DOI for MIT-BIH Supraventricular Arrhythmia Database (PhysioNet) (opens in new tab)"}]},{"slug":"mobilise-d-tvs","name":"Mobilise-D Technical Validation Study (TVS) dataset","tagline":null,"summary":"This dataset comes from the Mobilise-D Technical Validation Study, which develops and validates wearable tools for measuring gait and mobility in daily life. Participants included healthy people and patients with mobility-related conditions, recorded with wearable sensors in both lab and free-living settings.","signals":["imu"],"subjects":null,"license":"CC BY 4.0","gated":null,"updated":"2026-10-16","url":"/research/datasets/mobilise-d-tvs","links":[{"href":"https://zenodo.org/records/15861907","label":"Download","ariaLabel":"Download for Mobilise-D Technical Validation Study (TVS) dataset (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.15861907","label":"DOI","ariaLabel":"DOI for Mobilise-D Technical Validation Study (TVS) dataset (opens in new tab)"}]},{"slug":"moods-field-stress-fossil","name":"MOODS","tagline":"Mobile Open Observation of Daily Stressors (Fossil Sport WearOS watch, 100-day field study; private)","summary":"Neupane et al., \"Momentary Stressor Logging and Reflective Visualizations\", CHI '24 (DOI 10.1145/3613904.3642662; arXiv 2401.16307). Participants wore a Fossil Sport WearOS watch for 100 days; the watch triggered stress prompts when physiology suggested stress, and participants logged stressors and rated stress. IRB approved.","signals":["ppg","accel"],"subjects":122,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/moods-field-stress-fossil","links":[{"href":"https://doi.org/10.1145/3613904.3642662","label":"Download","ariaLabel":"Download for MOODS (opens in new tab)"}]},{"slug":"sleep-accel","name":"Motion and heart rate from a wrist-worn wearable and labeled sleep from polysomnography","tagline":null,"summary":null,"signals":["accel","hr"],"subjects":31,"license":"ODC-By 1.0","gated":false,"updated":"2026-10-14","url":"/research/datasets/sleep-accel","links":[{"href":"https://physionet.org/content/sleep-accel/1.0.0/","label":"Download","ariaLabel":"Download for Motion and heart rate from a wrist-worn wearable and labeled sleep from polysomnography (opens in new tab)"},{"href":"https://doi.org/10.13026/hmhs-py35","label":"DOI","ariaLabel":"DOI for Motion and heart rate from a wrist-worn wearable and labeled sleep from polysomnography (opens in new tab)"}]},{"slug":"mpmrd-multi-pace-running","name":"MPMRD","tagline":"Multi-Pace Multimodal Running Dataset (Huawei GT Runner watch + Polar Verity Sense + COSMED K5 + VICON)","summary":"Treadmill running from 30 healthy runners (8 female, 22 male). Each did five incremental relative-pace trials prescribed from individual VDOT, about 3 min per pace. Synchronized: VICON optical motion capture, COSMED PLUTO instrumented treadmill (kinetics), COSMED K5 portable metabolic analyzer, Polar Verity Sense optical HR sensor, HUAWEI GT Runner smartwatch, and Stryd Duo foot pods.","signals":["hr"],"subjects":"30 (8 F / 22 M healthy runners)","license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/mpmrd-multi-pace-running","links":[{"href":"https://zenodo.org/records/23098209","label":"Download","ariaLabel":"Download for MPMRD (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.23098209","label":"DOI","ariaLabel":"DOI for MPMRD (opens in new tab)"}]},{"slug":"mpu-rppg","name":"MPU-rPPG","tagline":"2K/60 fps multi-region (face, arms, hands, legs) rPPG video with CMS60D oximeter PPG (n=50, 20 h)","summary":"Chen, Tan, Yu, Jiang, Wang, Gao, Sun (Macao Polytechnic University, Great Bay University). Sci Data 13:1216 (2026), DOI 10.1038/s41597-026-07310-3, PMC13500808. Ethics: Zhuhai People's Hospital ZH20250306, written consent.","signals":["video","ppg","spo2"],"subjects":50,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/mpu-rppg","links":[{"href":"https://doi.org/10.6084/m9.figshare.29377835","label":"Download","ariaLabel":"Download for MPU-rPPG (opens in new tab)"},{"href":"https://doi.org/10.1038/s41597-026-07310-3","label":"DOI","ariaLabel":"DOI for MPU-rPPG (opens in new tab)"}]},{"slug":"mros","name":"MrOS Sleep Study (Osteoporotic Fractures in Men, NSRR)","tagline":null,"summary":"The MrOS Sleep Study is part of the Osteoporotic Fractures in Men Study and looks at how sleep disorders relate to falls, fractures, death and vascular disease. Of 5,994 men aged 65 or older in the parent study, 3,135 joined the sleep study between 2003 and 2005 and had unattended polysomnography and actigraphy. The NSRR release covers 2,911 men and focuses on the polysomnography data.","signals":["psg","ecg","eeg","resp","spo2"],"subjects":null,"license":"Not stated","gated":null,"updated":"2026-10-16","url":"/research/datasets/mros","links":[{"href":"https://sleepdata.org/datasets/mros","label":"Download","ariaLabel":"Download for MrOS Sleep Study (Osteoporotic Fractures in Men, NSRR) (opens in new tab)"},{"href":"https://doi.org/10.25822/kc27-0425","label":"DOI","ariaLabel":"DOI for MrOS Sleep Study (Osteoporotic Fractures in Men, NSRR) (opens in new tab)"}]},{"slug":"multisite-ppg","name":"Multi-site PPG (Earring / Necklace / Ring / Watch)","tagline":null,"summary":"The Multisite PPG dataset has multi-day free-living recordings from 20 participants wearing four custom devices at once: an earring, a necklace, a ring and a watch. Each device records green and infrared PPG, a 3-axis accelerometer and skin temperature at 100 Hz, and a Polar H10 chest strap supplies ECG at 130 Hz for heart-rate ground truth. It is about 18.6 GB, includes 8-second windows aligned with heart-rate labels, and is released under CC BY-NC 4.0.","signals":["ppg","ecg","accel","skin-temp"],"subjects":null,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/multisite-ppg","links":[{"href":"https://huggingface.co/datasets/snowballlab/Multisite-PPG","label":"Download","ariaLabel":"Download for Multi-site PPG (Earring / Necklace / Ring / Watch) (opens in new tab)"}]},{"slug":"multimodal-micromobility-stress","name":"Multimodal Micromobility Stress Dataset","tagline":"wrist BVP/EDA/temp/ACC on e-scooter rides","summary":"This dataset records riders during real-world micromobility trips on roads and sidewalks. For each participant it includes phone accelerometer data, eye tracking (gaze, fixations, blinks, saccades, pupil size), smartwatch signals (accelerometer, electrodermal activity, temperature, blood volume pulse) and frame-level stress labels from video feedback.","signals":["ppg","eda","skin-temp","accel"],"subjects":null,"license":"CC BY 4.0","gated":null,"updated":"2026-10-14","url":"/research/datasets/multimodal-micromobility-stress","links":[{"href":"https://figshare.com/articles/dataset/Multimodal_Micromobility_Stress_Dataset/31868863","label":"Download","ariaLabel":"Download for Multimodal Micromobility Stress Dataset (opens in new tab)"},{"href":"https://doi.org/10.6084/m9.figshare.31868863","label":"DOI","ariaLabel":"DOI for Multimodal Micromobility Stress Dataset (opens in new tab)"}]},{"slug":"muse-2-forehead-ppg-polar-validity","name":"Muse 2 forehead PPG vs Polar H10 concurrent validity (HR / ultra-short RMSSD)","tagline":null,"summary":"Griadhi, Suadnyana, Wahyuni (Universitas Dhyana Pura). Title on the record: Data and code for: Muse 2 forehead photoplethysmography agrees with Polar H10 for heart rate but overestimates ultra-short RMSSD. Ethics: Universitas Dhyana Pura 006821/KEP/2026.","signals":["ppg"],"subjects":28,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/muse-2-forehead-ppg-polar-validity","links":[{"href":"https://zenodo.org/records/23091944","label":"Download","ariaLabel":"Download for Muse 2 forehead PPG vs Polar H10 concurrent validity (HR / ultra-short RMSSD) (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.23091944","label":"DOI","ariaLabel":"DOI for Muse 2 forehead PPG vs Polar H10 concurrent validity (HR / ultra-short RMSSD) (opens in new tab)"}]},{"slug":"mw-ppg-hr","name":"MW-PPG-HR","tagline":"Multi-wavelength Wrist-worn PPG Dataset for Heart Rate Monitoring","summary":"Manchester Metropolitan University benchmark for PPG-based HR estimation, published Figshare 2026-03-05. Explicitly built to fix cohort-diversity and wavelength gaps in existing resources. Companion paper: \"Towards fair and trustworthy heart rate estimation from wrist-worn photoplethysmography: a multi-wavelength dataset and uncertainty-aware deep learning approach evaluated across skin tones, sexes and motion conditions.\"","signals":["ppg","ecg","accel"],"subjects":20,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/mw-ppg-hr","links":[{"href":"https://doi.org/10.6084/m9.figshare.30017974","label":"Download","ariaLabel":"Download for MW-PPG-HR (opens in new tab)"}]},{"slug":"mymove","name":"MyMove","tagline":"free-living older-adult smartwatch HAR (n=13, mean age 71)","summary":"MyMove (Kim / Choe / Kacorri et al., CHI 2022 DOI 10.1145/3491102.3517457; labeling pipeline IMWUT 2025). Northeast US in-the-wild deployment May–July 2021. Project page: https://mymove-collective.github.io/dataset/.","signals":["accel","magnetometer","hr"],"subjects":13,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/mymove","links":[{"href":"https://mymove-collective.github.io/dataset/","label":"Download","ariaLabel":"Download for MyMove (opens in new tab)"},{"href":"https://doi.org/10.1145/3491102.3517457","label":"DOI","ariaLabel":"DOI for MyMove (opens in new tab)"}]},{"slug":"neurobiosense","name":"NeuroBioSense","tagline":"neuromarketing Empatica E4 (Mendeley)","summary":"NeuroBioSense records physiological responses while participants watched 35 brand advertisements in three categories: cosmetics and fashion, cars and technology, and food and market. An Empatica E4 captured PPG, electrodermal activity and body temperature, and participants rated their emotions on an appraisal scale. It is aimed at neuromarketing research.","signals":["ppg","skin-temp","eda"],"subjects":58,"license":"CC BY 4.0","gated":null,"updated":"2026-10-14","url":"/research/datasets/neurobiosense","links":[{"href":"https://data.mendeley.com/datasets/7md7yty9sk/1","label":"Download","ariaLabel":"Download for NeuroBioSense (opens in new tab)"},{"href":"https://doi.org/10.17632/7md7yty9sk.1","label":"DOI","ariaLabel":"DOI for NeuroBioSense (opens in new tab)"}]},{"slug":"nfi-fared","name":"NFI_FARED","tagline":"forensic HAR, dominant-wrist + back IMU, controlled + free-living (n=14)","summary":"NFI_FARED is a forensic activity recognition dataset from the Netherlands Forensic Institute, released with the Hi-OSCAR paper on open-set activity classification. It contains accelerometer, gyroscope and air pressure data from 14 participants (8 men, 6 women, aged 26.6 ± 8.8) wearing one IMU on the dominant wrist and one on the back. Recordings are split into controlled sessions and unmonitored free-living periods, with labels such as walking, cycling, stairs and riding a train, car, tram or bus.","signals":["accel"],"subjects":14,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/nfi-fared","links":[{"href":"https://huggingface.co/datasets/NetherlandsForensicInstitute/NFI_FARED_IMU","label":"Download","ariaLabel":"Download for NFI_FARED (opens in new tab)"}]},{"slug":"nhanes-accelerometer","name":"NHANES 2011-2012 Physical Activity Monitor - Raw Data 80hz (PAX80_G)","tagline":null,"summary":"This is the raw 80 Hz accelerometer data from the 2011-2012 National Health and Nutrition Examination Survey (NHANES). Participants wore an ActiGraph GT3X+ for seven full days to capture 24-hour movement while awake and asleep. NCHS also releases the data summed per minute, hour and day in separate files.","signals":["accel"],"subjects":null,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/nhanes-accelerometer","links":[{"href":"https://wwwn.cdc.gov/Nchs/Data/Nhanes/Public/2011/DataFiles/PAX80_G.htm","label":"Download","ariaLabel":"Download for NHANES 2011-2012 Physical Activity Monitor - Raw Data 80hz (PAX80_G) (opens in new tab)"}]},{"slug":"nsr2db","name":"Normal Sinus Rhythm RR Interval Database (NSR2DB, PhysioNet)","tagline":null,"summary":"NSR2DB holds beat annotations for 54 long-term ECG recordings of people in normal sinus rhythm: 30 men aged 28.5 to 76 and 24 women aged 58 to 73. The original ECGs, which are not included, were digitized at 128 Hz, and the beats were annotated automatically and then reviewed by hand.","signals":["other"],"subjects":null,"license":"Not stated","gated":null,"updated":"2026-10-16","url":"/research/datasets/nsr2db","links":[{"href":"https://physionet.org/content/nsr2db/1.0.0/","label":"Download","ariaLabel":"Download for Normal Sinus Rhythm RR Interval Database (NSR2DB, PhysioNet) (opens in new tab)"},{"href":"https://doi.org/10.13026/C2S881","label":"DOI","ariaLabel":"DOI for Normal Sinus Rhythm RR Interval Database (NSR2DB, PhysioNet) (opens in new tab)"}]},{"slug":"numom2b","name":"nuMoM2b","tagline":"NSRR pregnancy sleep-disordered-breathing substudy","summary":"nuMoM2b (Nulliparous Pregnancy Outcomes Study: Monitoring Mothers-to-Be) is a prospective cohort of women pregnant for the first time, studying the causes of common adverse pregnancy outcomes. Of 10,038 women, 3,702 took part in a sleep-disordered breathing substudy between 2011 and 2013, with two visits during pregnancy. The sleep data are shared through the National Sleep Research Resource.","signals":["psg","ppg"],"subjects":3702,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/numom2b","links":[{"href":"https://sleepdata.org/datasets/numom2b","label":"Download","ariaLabel":"Download for nuMoM2b (opens in new tab)"}]},{"slug":"nurse-hospital-stress","name":"NURSE (hospital stress, E4, Dryad)","tagline":null,"summary":"Multi-modal continuous stress detection of nurses in a real hospital during COVID-19. Dryad DOI 10.5061/dryad.5hqbzkh6f. ~1,250 hours across two 2020 sessions.","signals":["ppg","accel","skin-temp","eda"],"subjects":15,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/nurse-hospital-stress","links":[{"href":"https://datadryad.org/dataset/doi:10.5061/dryad.5hqbzkh6f","label":"Download","ariaLabel":"Download for NURSE (hospital stress, E4, Dryad) (opens in new tab)"},{"href":"https://doi.org/10.5061/dryad.5hqbzkh6f","label":"DOI","ariaLabel":"DOI for NURSE (hospital stress, E4, Dryad) (opens in new tab)"}]},{"slug":"office-everyday-mwl","name":"Office Everyday MWL (MW_OFFICE, n=10 E4)","tagline":null,"summary":"This dataset was recorded from 10 participants doing relaxation, reading, summarizing and mental workload tasks with consumer-grade wearables. Each person has two controlled lab sessions and one in-the-wild session. It suits work on real-time cognitive load detection and on cleaning artifacts from low-cost wearable data.","signals":["ppg","accel","skin-temp","eda","eeg"],"subjects":10,"license":"CC BY 4.0","gated":null,"updated":"2026-10-14","url":"/research/datasets/office-everyday-mwl","links":[{"href":"https://zenodo.org/records/15681263","label":"Download","ariaLabel":"Download for Office Everyday MWL (MW_OFFICE, n=10 E4) (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.15681263","label":"DOI","ariaLabel":"DOI for Office Everyday MWL (MW_OFFICE, n=10 E4) (opens in new tab)"}]},{"slug":"openneuro-ds007537","name":"OpenNeuro ds007537","tagline":"smartphone interactions multimodal (PPG+EEG+ET)","summary":"This OpenNeuro dataset records EEG, eye tracking, PPG and galvanic skin response from 23 healthy adults during natural smartphone use and while watching a video. Each person used their own phone freely for 10 minutes and watched a standard video for 5 minutes, wearing a 64-channel EEG cap and Tobii Pro Glasses 2. Hardware trigger pulses align all streams, and the data are in BIDS format under CC0.","signals":["ppg","eda","eeg","eye-tracking"],"subjects":23,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/openneuro-ds007537","links":[{"href":"https://openneuro.org/datasets/ds007537","label":"Download","ariaLabel":"Download for OpenNeuro ds007537 (opens in new tab)"},{"href":"https://doi.org/10.18112/openneuro.ds007537.v1.0.0","label":"DOI","ariaLabel":"DOI for OpenNeuro ds007537 (opens in new tab)"}]},{"slug":"openoximetry","name":"OpenOximetry Repository (PhysioNet Restricted)","tagline":null,"summary":"The OpenOximetry Repository collects clinical and lab pulse oximetry data so researchers can study how physiology and device factors affect oximeter accuracy, including skin color. The first release holds controlled desaturation studies with paired oximeter and arterial blood gas readings, raw PPG, ECG, arterial pressure, EtCO2 and skin color measurements. It is run by the UCSF Hypoxia Lab and partners and is hosted on PhysioNet with restricted access.","signals":["ppg","ecg","spo2","abp","etco2"],"subjects":null,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/openoximetry","links":[{"href":"https://physionet.org/content/openox-repo/1.1.1/","label":"Download","ariaLabel":"Download for OpenOximetry Repository (PhysioNet Restricted) (opens in new tab)"}]},{"slug":"openwatch-smartwatch-gesture-imu-ppg","name":"OpenWatch","tagline":"smartwatch IMU+PPG hand-gesture benchmark (n=50, >10 h, 59 gesture sequences)","summary":"OpenWatch is an open benchmark for wrist gesture recognition using synchronized IMU and PPG data from a commercial smartwatch, with a subject-independent evaluation protocol. The authors also propose MixToken, a small task-specific mixture-of-experts model, and NormWear-LoRA, a low-rank adaptation of a smartwatch foundation model. Adding PPG improved F1 by 12.5% for foundation models, and MixToken beat fine-tuned foundation models (F1 90% vs 66%) with far fewer parameters (223k vs 136M).","signals":["accel","gyro","ppg"],"subjects":50,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/openwatch-smartwatch-gesture-imu-ppg","links":[{"href":"https://arxiv.org/abs/2605.04791","label":"Download","ariaLabel":"Download for OpenWatch (opens in new tab)"},{"href":"https://doi.org/10.48550/arXiv.2605.04791","label":"DOI","ariaLabel":"DOI for OpenWatch (opens in new tab)"}]},{"slug":"opportunity","name":"opportunity","tagline":null,"summary":"OPPORTUNITY is a benchmark for activity recognition from body-worn, object and ambient sensors, recorded while users carried out daily activities. Body sensors include 7 IMUs and 12 3D accelerometers, alongside sensors on 12 objects and in the room. Four users each did 6 runs, five natural daily-living runs and one scripted drill, with labels at several levels from locomotion to fine actions; a subset was used for the 2011 OPPORTUNITY challenge.","signals":["accel"],"subjects":4,"license":"Creative Commons Attribution 4.0 International (CC BY 4.0)","gated":false,"updated":"2026-10-14","url":"/research/datasets/opportunity","links":[{"href":"https://archive.ics.uci.edu/dataset/226/opportunity+activity+recognition","label":"Download","ariaLabel":"Download for opportunity (opens in new tab)"},{"href":"https://doi.org/10.24432/C5M027","label":"DOI","ariaLabel":"DOI for opportunity (opens in new tab)"}]},{"slug":"osasud","name":"OSASUD","tagline":"Obstructive Sleep Apnea Stroke Unit Dataset (finger PPG+ECG+SpO₂)","summary":"OSASUD (Bernardini / Brunello / Gigli / Montanari / Saccomanno; Udine University Hospital stroke unit + University of Udine). Simultaneous overnight Mindray vital-signs monitoring and Embletta level-3 PSG (Aug 2019–Jul 2020). PSG scored by a trained sleep physician (AASM rules); annotations temporally aligned to Mindray.","signals":["ppg","ecg","spo2","hr","resp","psg"],"subjects":30,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/osasud","links":[{"href":"https://doi.org/10.6084/m9.figshare.c.5630890","label":"Download","ariaLabel":"Download for OSASUD (opens in new tab)"}]},{"slug":"paaws","name":"PAAWS R1","tagline":"multi-site accelerometer/IMU, lab + 7-day free-living + home PSG (n=252)","summary":"PAAWS R1 (Dec 2025, Northeastern / Intille lab; IMWUT 9(4) art. 204, doi:10.1145/3770639): ~4 h SimFL+Lab per person from 252 individuals with accelerometer/IMU at up to 21 locations (wrists top/bottom, waist, thighs, ankles) plus heart rate and metabolic data; ~7 days 24-h free-living from 20 adults (wrists, waist, thigh, ankle, phone); 29 PSG nights at home. Second-by-second video-annotated labels.","signals":["accel"],"subjects":252,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/paaws","links":[{"href":"https://www.paawsstudy.org/data.html","label":"Download","ariaLabel":"Download for PAAWS R1 (opens in new tab)"}]},{"slug":"pamap2","name":"PAMAP2","tagline":"Physical Activity Monitoring (n=9, 3× IMU + HR)","summary":"PAMAP2 (Physical Activity Monitoring for Aging People 2) released via UCI ML Repository (DOI 10.24432/C5NW2H). Protocol + optional sessions as .dat files (54 columns). Introductory papers: Reiss & Stricker ISWC 2012 / ABRA 2012.","signals":["accel","gyro","magnetometer","skin-temp","hr"],"subjects":9,"license":"CC BY 4.0","gated":false,"updated":"2026-10-15","url":"/research/datasets/pamap2","links":[{"href":"https://archive.ics.uci.edu/dataset/231/pamap2+physical+activity+monitoring","label":"Download","ariaLabel":"Download for PAMAP2 (opens in new tab)"},{"href":"https://doi.org/10.24432/C5NW2H","label":"DOI","ariaLabel":"DOI for PAMAP2 (opens in new tab)"}]},{"slug":"parkinsensedb","name":"ParkinSenseDB","tagline":"wearable IMU+EMG + optoelectronic gait (PD vs control, n=53)","summary":"ParkinSenseDB (Raggiunto / Belli / Paniccia et al.). Mendeley Data version 2, published 2026-09-12. DOI 10.17632/brpx8jwsh7.2.","signals":["accel","gyro","magnetometer"],"subjects":53,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/parkinsensedb","links":[{"href":"https://data.mendeley.com/datasets/brpx8jwsh7/2","label":"Download","ariaLabel":"Download for ParkinSenseDB (opens in new tab)"},{"href":"https://doi.org/10.17632/brpx8jwsh7.2","label":"DOI","ariaLabel":"DOI for ParkinSenseDB (opens in new tab)"}]},{"slug":"ppg-device-appel","name":"Photoplethysmography Device Data (Appel & Theart / Mendeley)","tagline":null,"summary":"This dataset accompanies a study on designing a PPG device for clinical use by Anneri Appel (Aspeling) and colleagues. It has raw and processed data from 110 participants (one excluded for a measurement error), recorded with both green-LED and red-LED devices, plus per-participant metadata such as age, sex and BMI class.","signals":["ppg"],"subjects":109,"license":"CC BY-NC 3.0","gated":null,"updated":"2026-10-15","url":"/research/datasets/ppg-device-appel","links":[{"href":"https://data.mendeley.com/datasets/bcb9329f6d/1","label":"Download","ariaLabel":"Download for Photoplethysmography Device Data (Appel & Theart / Mendeley) (opens in new tab)"},{"href":"https://doi.org/10.17632/bcb9329f6d.1","label":"DOI","ariaLabel":"DOI for Photoplethysmography Device Data (Appel & Theart / Mendeley) (opens in new tab)"}]},{"slug":"challenge-2017","name":"PhysioNet/CinC Challenge 2017: AF classification from a short single-lead ECG (AliveCor)","tagline":null,"summary":"The PhysioNet/Computing in Cardiology Challenge 2017 dataset contains single-lead ECGs recorded by patients on AliveCor handheld devices. The public training set has 8,528 recordings lasting 9 to just over 60 seconds, sampled at 300 Hz, each hand-labelled as normal, atrial fibrillation, other rhythm or too noisy. A further 3,658 recordings form a private test set.","signals":["ecg"],"subjects":null,"license":"Not stated","gated":null,"updated":"2026-10-16","url":"/research/datasets/challenge-2017","links":[{"href":"https://physionet.org/content/challenge-2017/1.0.0/","label":"Download","ariaLabel":"Download for PhysioNet/CinC Challenge 2017: AF classification from a short single-lead ECG (AliveCor) (opens in new tab)"},{"href":"https://doi.org/10.13026/d3hm-sf11","label":"DOI","ariaLabel":"DOI for PhysioNet/CinC Challenge 2017: AF classification from a short single-lead ECG (AliveCor) (opens in new tab)"}]},{"slug":"paee-utwente-cosmed-imu-ecg","name":"PM-EKF energy expenditure","tagline":"3x Xsens DOT IMU (thighs + pelvis) + 1-lead ECG + COSMED K5 indirect calorimetry (10 subj, U Twente eHealth House)","summary":"This dataset was used to develop PM-EKF, a physiological model-based extended Kalman filter for estimating physical activity energy expenditure in daily life. It covers 10 participants (30% female), with age, sex, height and weight recorded and body composition measured on an Omron BF511 bioimpedance scale.","signals":["accel","ecg"],"subjects":10,"license":"CC BY 4.0","gated":null,"updated":"2026-10-15","url":"/research/datasets/paee-utwente-cosmed-imu-ecg","links":[{"href":"https://zenodo.org/records/19388845","label":"Download","ariaLabel":"Download for PM-EKF energy expenditure (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.19388845","label":"DOI","ariaLabel":"DOI for PM-EKF energy expenditure (opens in new tab)"}]},{"slug":"podcast-ring-ppg","name":"Podcast playback-speed physio","tagline":"EEG+EDA+PPG (n=62, Zenodo 22541594)","summary":"Scientific Data companion deposit (An / Pak / Désiron / Schneider / Donati; ETH Zurich / UZH / Fribourg). Version 1.0 zip playbackspeeddatasetv1.0.0.zip. English cohort includes workload/engagement/comprehension self-reports; raw EEG for 61/62.","signals":["ppg","eda","eeg"],"subjects":62,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/podcast-ring-ppg","links":[{"href":"https://zenodo.org/records/22541594","label":"Download","ariaLabel":"Download for Podcast playback-speed physio (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.22541594","label":"DOI","ariaLabel":"DOI for Podcast playback-speed physio (opens in new tab)"}]},{"slug":"ppg-cognitive-fatigue-5gamers","name":"PPG Heart Beat for Cognitive Fatigue Prediction (5-Gamers)","tagline":null,"summary":"Kaggle canaria/5-gamers. Long continuous recordings, sparse hourly annotations (SSS, sleep-2-peak RT, caffeine/food). GitHub: canaria-technologies/kaggle-5-gamers-input.","signals":["ppg"],"subjects":5,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/ppg-cognitive-fatigue-5gamers","links":[{"href":"https://www.kaggle.com/datasets/canaria/5-gamers","label":"Download","ariaLabel":"Download for PPG Heart Beat for Cognitive Fatigue Prediction (5-Gamers) (opens in new tab)"}]},{"slug":"ppg-bgl-kermani","name":"PPG Signals for Blood Glucose Measurement (Kermani / Mendeley)","tagline":null,"summary":"This dataset holds 67 raw PPG recordings sampled at 2,175 Hz, each labelled with age, gender and an invasive blood glucose reading. It was recorded by a digital systems research team at the University of Science and Technology of Mazandaran in Behshahr, Iran, for work on non-invasive glucose estimation.","signals":["ppg"],"subjects":23,"license":"CC BY 4.0","gated":false,"updated":"2026-10-14","url":"/research/datasets/ppg-bgl-kermani","links":[{"href":"https://data.mendeley.com/datasets/37pm7jk7jn/3","label":"Download","ariaLabel":"Download for PPG Signals for Blood Glucose Measurement (Kermani / Mendeley) (opens in new tab)"},{"href":"https://doi.org/10.17632/37pm7jk7jn.3","label":"DOI","ariaLabel":"DOI for PPG Signals for Blood Glucose Measurement (Kermani / Mendeley) (opens in new tab)"}]},{"slug":"ppg-bgl-smartwatch-bhatia","name":"PPG-BGL Smartwatch (Bhatia & Ekbote 2026 / Zenodo 20577959)","tagline":null,"summary":"This dataset pairs continuous smartwatch PPG with glucose values from a continuous glucose monitor (CGM), so models can be trained to estimate glucose from PPG. The authors also report early experiments suggesting reasonable accuracy may be possible, but say more data from more people is needed.","signals":["ppg","cgm"],"subjects":1,"license":"CC BY 4.0","gated":null,"updated":"2026-10-14","url":"/research/datasets/ppg-bgl-smartwatch-bhatia","links":[{"href":"https://zenodo.org/records/20577959","label":"Download","ariaLabel":"Download for PPG-BGL Smartwatch (Bhatia & Ekbote 2026 / Zenodo 20577959) (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.20577959","label":"DOI","ariaLabel":"DOI for PPG-BGL Smartwatch (Bhatia & Ekbote 2026 / Zenodo 20577959) (opens in new tab)"}]},{"slug":"ppgbp-guilin","name":"PPG-BP Database (Guilin / Liang 2018)","tagline":null,"summary":"PPG-BP Database (Liang et al. lineage; Guilin). Fingertip reflective PPG with one cuff BP reading per subject and hypertension-stage labels.","signals":["ppg"],"subjects":219,"license":"CC0 1.0","gated":null,"updated":"2026-10-14","url":"/research/datasets/ppgbp-guilin","links":[{"href":"https://figshare.com/articles/dataset/PPG-BP_Database/5459299","label":"Download","ariaLabel":"Download for PPG-BP Database (Guilin / Liang 2018) (opens in new tab)"}]},{"slug":"ppg-dalia","name":"PPG-DaLiA","tagline":"PPG dataset for motion compensation / HR","summary":"PPG-DaLiA (Reiss / Indlekofer / Schmidt / Van Laerhoven lineage; Uni-Siegen). Daily-life activities with wrist PPG+accel and chest physiological reference. USI page: https://ubi29.informatik.uni-siegen.de/usi/datappgdalia.html.","signals":["ppg","ecg","accel","skin-temp","resp"],"subjects":15,"license":"Scientific, non-commercial use with credit to the owners","gated":null,"updated":"2026-10-15","url":"/research/datasets/ppg-dalia","links":[{"href":"https://uni-siegen.sciebo.de/s/pfHzlTepXkiJ4jP","label":"Download","ariaLabel":"Download for PPG-DaLiA (opens in new tab)"}]},{"slug":"ppg-signal-dataset-usc-cali","name":"PPG-Signal-Dataset (USC Cali)","tagline":"56 x 2-min PPG with single SBP/DBP + JNC/AHA class","summary":"This study tested whether non-fiducial PPG features can estimate blood pressure without relying on detecting specific points in the waveform. The authors recorded PPG from 56 participants, computed 57 non-fiducial features such as statistical indexes and energy operators, and combined feature selection with machine learning. Gaussian process regression with ReliefF-selected features gave the best systolic and diastolic predictions.","signals":["ppg"],"subjects":56,"license":"CC BY 4.0","gated":null,"updated":"2026-10-14","url":"/research/datasets/ppg-signal-dataset-usc-cali","links":[{"href":"https://github.com/sanvsquezsz/PPG-based-BP-assessment","label":"Download","ariaLabel":"Download for PPG-Signal-Dataset (USC Cali) (opens in new tab)"},{"href":"https://doi.org/10.1007/s13246-025-01593-3","label":"DOI","ariaLabel":"DOI for PPG-Signal-Dataset (USC Cali) (opens in new tab)"}]},{"slug":"ppg-sport","name":"PPG-Sport","tagline":"dual-wrist EmbracePlus sports HR (ECG-ref)","summary":"PPG-Sport (Hu / Pham / … / Ma), Proc. ACM IMWUT 2026, DOI 10.1145/3810216. First large public wrist-PPG sports set targeting non-periodic, high-intensity, bilaterally asymmetric motion.","signals":["ppg","ecg","accel","gyro"],"subjects":30,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/ppg-sport","links":[{"href":"https://github.com/LaserHu/PPG-Sport","label":"Download","ariaLabel":"Download for PPG-Sport (opens in new tab)"},{"href":"https://doi.org/10.1145/3810216","label":"DOI","ariaLabel":"DOI for PPG-Sport (opens in new tab)"}]},{"slug":"ppgarrhythmia","name":"PPGArrhythmia (Liu 2022)","tagline":"public valid/test subset only","summary":"PulseLM cites full 118,217 segments / 228 patients. Only validtestDataset/ (~91 .mat files) is on GitHub. Full train set requires author access.","signals":["ppg"],"subjects":91,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/ppgarrhythmia","links":[{"href":"https://github.com/zdzdliu/PPGArrhythmiaDetection","label":"Download","ariaLabel":"Download for PPGArrhythmia (Liu 2022) (opens in new tab)"},{"href":"https://doi.org/10.1161/JAHA.121.023555","label":"DOI","ariaLabel":"DOI for PPGArrhythmia (Liu 2022) (opens in new tab)"}]},{"slug":"ptb-xl","name":"PTB-XL","tagline":"21,799 clinical 12-lead ECGs (PhysioNet)","summary":"21,799 10-s 12-lead ECGs from 18,869 patients (Schiller devices, 1989–1996), annotated with 71 SCP-ECG statements (diagnostic, form, rhythm — rhythm statements include AF), with recommended 10-fold patient-wise splits (folds 9–10 human-validated). v1.0.3, 3.0 GB, open access. Source: PhysioNet page (fetched 2026-10-12).","signals":["ecg"],"subjects":18869,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/ptb-xl","links":[{"href":"https://physionet.org/content/ptb-xl/1.0.3/","label":"Download","ariaLabel":"Download for PTB-XL (opens in new tab)"},{"href":"https://doi.org/10.13026/kfzx-aw45","label":"DOI","ariaLabel":"DOI for PTB-XL (opens in new tab)"}]},{"slug":"pulse-transit-time-ppg","name":"Pulse Transit Time PPG","tagline":"multi-λ finger PPG+ECG+IMU (n=22, PhysioNet)","summary":"Pulse Transit Time PPG Dataset v1.1.0 (PhysioNet). High-resolution synchronized multi-sensor finger PPG under three activities.","signals":["ppg","ecg","accel","gyro"],"subjects":22,"license":"ODbL 1.0","gated":false,"updated":"2026-10-14","url":"/research/datasets/pulse-transit-time-ppg","links":[{"href":"https://physionet.org/content/pulse-transit-time-ppg/1.1.0/","label":"Download","ariaLabel":"Download for Pulse Transit Time PPG (opens in new tab)"},{"href":"https://doi.org/10.13026/jpan-6n92","label":"DOI","ariaLabel":"DOI for Pulse Transit Time PPG (opens in new tab)"}]},{"slug":"pulse-wave-database","name":"Pulse Wave Database (PWDB)","tagline":"simulated arterial pulse waves incl. finger/wrist-site PPG, 4,374 virtual adults","summary":null,"signals":["ppg","abp"],"subjects":"4374 virtual subjects (in silico; ages 25–75 in 10-year bands)","license":"ODC-PDDL","gated":false,"updated":"2026-10-14","url":"/research/datasets/pulse-wave-database","links":[{"href":"https://zenodo.org/records/3275625","label":"Download","ariaLabel":"Download for Pulse Wave Database (PWDB) (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.3275625","label":"DOI","ariaLabel":"DOI for Pulse Wave Database (PWDB) (opens in new tab)"}]},{"slug":"pulsedb","name":"PulseDB v2.0","tagline":"cleaned MIMIC-III + VitalDB cuffless-BP benchmark (MIMIC part ODbL; VitalDB part CC BY-NC-SA)","summary":"Wang et al., PulseDB: A large, cleaned dataset based on MIMIC-III and VitalDB for benchmarking cuff-less blood pressure estimation methods, Frontiers in Digital Health (2023), doi:10.3389/fdgth.2022.1090854. Repo: https://github.com/pulselabteam/PulseDB (@db0824f, v20 on main). The v2.0 release holds 5,361 subject files of 10 s ECG/PPG/ABP segments (README).","signals":["ppg","ecg","abp"],"subjects":5361,"license":"Split by folder: PulseDB_MIMIC files under ODbL 1.0 (contents under DbCL 1.0); PulseDB_Vital files and the subset/Info files under CC BY-NC-SA 4.0","gated":null,"updated":"2026-10-15","url":"/research/datasets/pulsedb","links":[{"href":"https://github.com/pulselabteam/PulseDB","label":"Download","ariaLabel":"Download for PulseDB v2.0 (opens in new tab)"},{"href":"https://doi.org/10.3389/fdgth.2022.1090854","label":"DOI","ariaLabel":"DOI for PulseDB v2.0 (opens in new tab)"}]},{"slug":"pulsewatch","name":"Pulsewatch","tagline":"ambulatory smartwatch PPG with AF adjudication (Synapse gated)","summary":"Pulsewatch is a Synapse-hosted clinical/ambulatory AF monitoring dataset and companion code release (UConn / PulsewatchRelease). Companion ICASSP 2025 training/eval code documents 30 s PPG/HR/ACC products and AF adjudication labels.","signals":["ppg","ecg","accel"],"subjects":106,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/pulsewatch","links":[{"href":"https://www.synapse.org/Synapse:syn23565056","label":"Download","ariaLabel":"Download for Pulsewatch (opens in new tab)"}]},{"slug":"ppg-bp-annot-zenodo","name":"pyPPG benchmarking","tagline":"PPG-BP fiducial annotations (Zenodo 10523285)","summary":"Zenodo record 10523285 (publication date 2024-01-17; pinned by record id + file md5, no metadata.version string). Title: pyPPG benchmarking. Acquired archive: PPG-BPannot.zip only (results.zip deliberately not downloaded).","signals":[],"subjects":219,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/ppg-bp-annot-zenodo","links":[{"href":"https://zenodo.org/records/10523285","label":"Download","ariaLabel":"Download for pyPPG benchmarking (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.10523285","label":"DOI","ariaLabel":"DOI for pyPPG benchmarking (opens in new tab)"}]},{"slug":"qol-stress","name":"QoL_Stress","tagline":"ECG/EDA + Fitbit HR with self-reported stress","summary":"QoLStress is a multimodal physiological + self-report stress dataset for stress detection research. Zenodo record 20757481 (created 2026-06-22), license CC BY 4.0. Collections: ecg, eda, Fitbit HR streams, demographics, PSS-14, STAI-S.","signals":["ecg","eda","fitbit-hr"],"subjects":66,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/qol-stress","links":[{"href":"https://zenodo.org/records/20757481","label":"Download","ariaLabel":"Download for QoL_Stress (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.20757481","label":"DOI","ariaLabel":"DOI for QoL_Stress (opens in new tab)"}]},{"slug":"radboud-ptt-ppg-bp","name":"Radboud PTT PPG+BP (Mol 2020)","tagline":"wrist+finger PPG with Finapres BP","summary":"Official title: Pulse transit time as a proxy for vasoconstriction in younger and older adults (version 1). Creators: Arjen Mol, Carel G.M. Meskers, Sjoerd P.","signals":["ppg","ecg","bp-finapres","tpr","nirs","spo2","resp"],"subjects":86,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/radboud-ptt-ppg-bp","links":[{"href":"https://data.ru.nl/collections/di/dcn/DSC_626810_0007_173","label":"Download","ariaLabel":"Download for Radboud PTT PPG+BP (Mol 2020) (opens in new tab)"},{"href":"https://doi.org/10.34973/te70-x603","label":"DOI","ariaLabel":"DOI for Radboud PTT PPG+BP (Mol 2020) (opens in new tab)"}]},{"slug":"reallifehar","name":"Real-Life HAR (UDC)","tagline":"smartphone acc/gyro/mag/GPS, real-life (n=19)","summary":"This dataset from the Universidade da Coruña database lab records smartphone sensors during real everyday use by 19 people. It includes accelerometer, gyroscope, magnetometer and GPS data, each labelled as inactive, active, walking or driving, and is meant to be independent of phone orientation, placement and user. The authors also report preliminary SVM experiments.","signals":["accel"],"subjects":19,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/reallifehar","links":[{"href":"https://lbd.udc.es/research/real-life-HAR-dataset/","label":"Download","ariaLabel":"Download for Real-Life HAR (UDC) (opens in new tab)"}]},{"slug":"realworld-ppg-siam","name":"Real-World PPG Dataset (Siam et al. / Mendeley)","tagline":null,"summary":"This dataset contains short real-world PPG recordings of 300 samples each (6 seconds at 50 Hz). It comes pre-split into a training file of 1,374 signals (about 66%) and a test file of 700 signals (about 34%).","signals":["ppg"],"subjects":35,"license":"CERN OHL version 1.2","gated":null,"updated":"2026-10-14","url":"/research/datasets/realworld-ppg-siam","links":[{"href":"https://data.mendeley.com/datasets/yynb8t9x3d/3","label":"Download","ariaLabel":"Download for Real-World PPG Dataset (Siam et al. / Mendeley) (opens in new tab)"},{"href":"https://doi.org/10.17632/yynb8t9x3d.3","label":"DOI","ariaLabel":"DOI for Real-World PPG Dataset (Siam et al. / Mendeley) (opens in new tab)"}]},{"slug":"realdisp","name":"REALDISP","tagline":"9-site IMU fitness HAR with sensor displacement (n=17)","summary":"REALDISP (REAListic sensor DISPlacement) was collected to study how moving wearable sensors affects activity recognition and to benchmark recognition methods in general. Seventeen subjects did warm-up, cool-down and fitness exercises with sensors in three conditions: ideal placement, self-placement by the user and deliberately induced displacement. Sensors recorded acceleration, rate of turn, magnetic field and orientation quaternions.","signals":["accel"],"subjects":17,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/realdisp","links":[{"href":"https://archive.ics.uci.edu/dataset/305/realdisp+activity+recognition+dataset","label":"Download","ariaLabel":"Download for REALDISP (opens in new tab)"}]},{"slug":"realworld-har","name":"RealWorld HAR","tagline":"multi-position smartphone/smartwatch IMU (n=15)","summary":"Introduced with On-body localization of wearable devices… (IEEE PerCom 2016; Sztyler & Stuckenschmidt, University of Mannheim DWS). Subjects performed climbing stairs down/up, jumping, lying, standing, sitting, running/jogging, and walking while seven on-body devices recorded sensors concurrently. Official landing: Mannheim DWS RealWorld dataset page.","signals":["accel","gyro","magnetometer","audio","gps","light"],"subjects":15,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/realworld-har","links":[{"href":"https://www.uni-mannheim.de/dws/research/projects/activity-recognition/dataset/dataset-realworld/","label":"Download","ariaLabel":"Download for RealWorld HAR (opens in new tab)"}]},{"slug":"recgym","name":"RecGym","tagline":"gym workouts, IMU + body capacitance at wrist/pocket/calf (n=10, 5 days)","summary":"RecGym records gym workouts from ten volunteers using a sensing unit with an IMU and a body capacitance sensor, worn at the wrist, in the pocket and on the calf at 20 Hz. It covers eleven workouts, such as bench press, squats and rope skipping, plus a null class. Each volunteer did five roughly one-hour sessions on five days, for fifty sessions in total.","signals":["accel"],"subjects":10,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/recgym","links":[{"href":"https://zhaxidele.github.io/RecGym/","label":"Download","ariaLabel":"Download for RecGym (opens in new tab)"}]},{"slug":"brain-wearable-monitoring","name":"Regulation of Brain Cognitive States through Auditory, Gustatory, and Olfactory Stimulation with Wearable Monitoring","tagline":null,"summary":"This PhysioNet dataset comes from experiments that tried to shift people's cognitive arousal and performance with safe sensory interventions. In one experiment participants listened to different genres of music; in the other they drank coffee and smelled perfume. Physiological data were recorded with two Empatica wristbands and a Muse EEG headband.","signals":["ppg","accel","eda","eeg"],"subjects":null,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/brain-wearable-monitoring","links":[{"href":"https://physionet.org/content/brain-wearable-monitoring/1.0.0/","label":"Download","ariaLabel":"Download for Regulation of Brain Cognitive States through Auditory, Gustatory, and Olfactory Stimulation with Wearable Monitoring (opens in new tab)"},{"href":"https://doi.org/10.13026/jr6f-wc85","label":"DOI","ariaLabel":"DOI for Regulation of Brain Cognitive States through Auditory, Gustatory, and Olfactory Stimulation with Wearable Monitoring (opens in new tab)"}]},{"slug":"respiratory-oximetry-apnoea","name":"Respiratory and Pulse Oximetry Waveforms from Healthy Adults During Simulated Apnoea Events","tagline":null,"summary":"This dataset records airway pressure, inspiratory differential pressure and pulse oximetry from 20 healthy adults (10 women, 10 men) during simulated obstructive sleep apnoea. A custom neck-worn reflectance oximeter captured PPG from both the carotid artery and the jugular vein. Each trial started with normal breathing, then 10- and 20-second breath holds, and it is meant for OSA detection and for estimating arterial and venous oxygen saturation.","signals":["ppg","resp"],"subjects":null,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/respiratory-oximetry-apnoea","links":[{"href":"https://physionet.org/content/respiratory-oximetry-apnoea/1.0.0/","label":"Download","ariaLabel":"Download for Respiratory and Pulse Oximetry Waveforms from Healthy Adults During Simulated Apnoea Events (opens in new tab)"},{"href":"https://doi.org/10.13026/s45r-k263","label":"DOI","ariaLabel":"DOI for Respiratory and Pulse Oximetry Waveforms from Healthy Adults During Simulated Apnoea Events (opens in new tab)"}]},{"slug":"rest-asmr-ppg","name":"REST-ASMR PPG Dataset (Zenodo)","tagline":null,"summary":"REST-ASMR records behavioural reports and PPG while people watched ASMR videos, with nature-relaxation videos as a control. The stimuli worked on 97% of participants, and PPG showed a heart-rate slowdown specific to ASMR. A bidirectional LSTM separated ASMR from nature videos perfectly at the video level and reached 75.51% frame-level accuracy for tingle states under subject- and video-independent cross-validation.","signals":["ppg"],"subjects":34,"license":"CC BY 4.0","gated":null,"updated":"2026-10-14","url":"/research/datasets/rest-asmr-ppg","links":[{"href":"https://zenodo.org/records/18881334","label":"Download","ariaLabel":"Download for REST-ASMR PPG Dataset (Zenodo) (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.18881334","label":"DOI","ariaLabel":"DOI for REST-ASMR PPG Dataset (Zenodo) (opens in new tab)"}]},{"slug":"samsung-afib-pvc-patch-study","name":"Samsung AFib/PVC study","tagline":"Galaxy Watch PPG with Preventice ECG-patch beat labels (private)","summary":"Samsung study pairing watch PPG with an ambulatory ECG patch. xMAE describes it as a lab setting; MMR's Table 4 labels it \"Naturalistic (field)\". The setting is therefore reported inconsistently.","signals":["ppg"],"subjects":139,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/samsung-afib-pvc-patch-study","links":[{"href":"https://doi.org/10.48550/arXiv.2605.00973","label":"Download","ariaLabel":"Download for Samsung AFib/PVC study (opens in new tab)"}]},{"slug":"samsung-bp-free-living","name":"Samsung BP free-living study","tagline":"Galaxy Watch PPG with cuff-based BP (private)","summary":"Free-living Samsung study pairing watch PPG with cuff BP readings. HiMAE calls it \"My BP Lab (Clinical Trials ID 19-27169)\", free-world setting.","signals":["ppg"],"subjects":9427,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/samsung-bp-free-living","links":[{"href":"https://doi.org/10.48550/arXiv.2605.00973","label":"Download","ariaLabel":"Download for Samsung BP free-living study (opens in new tab)"}]},{"slug":"samsung-bp-lab-cnap","name":"Samsung BP lab study","tagline":"Galaxy Watch PPG with CNAP continuous finger BP reference (private)","summary":"Samsung lab study pairing resting wrist PPG with CNAP beat-to-beat finger BP. HiMAE calls it the \"My Heart Lab Study (NCT04314947)\". Conflict: ClinicalTrials.gov lists NCT04314947's org study ID as \"MyHeartLab Free Living Study\" (record details withheld because the device isn't FDA-cleared), so the NCT mapping to this lab set is unconfirmed.","signals":["ppg"],"subjects":135,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/samsung-bp-lab-cnap","links":[{"href":"https://doi.org/10.48550/arXiv.2605.00973","label":"Download","ariaLabel":"Download for Samsung BP lab study (opens in new tab)"}]},{"slug":"samsung-tulane-lab-tests","name":"Samsung Galaxy Watch PPG + clinical lab panel (Tulane University study 20242033; private)","tagline":null,"summary":"Watch PPG paired with clinical lab values. Used as \"lab tests\" regression/classification targets.","signals":["ppg"],"subjects":45,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/samsung-tulane-lab-tests","links":[{"href":"https://doi.org/10.48550/arXiv.2510.25785","label":"Download","ariaLabel":"Download for Samsung Galaxy Watch PPG + clinical lab panel (Tulane University study 20242033; private) (opens in new tab)"}]},{"slug":"samsung-galaxy-watch-pretraining-corpus","name":"Samsung Galaxy Watch PPG pretraining corpus (~32K users, ~17M 10-s segments; private)","tagline":null,"summary":"Free-living wrist PPG pooled from Samsung research studies. Not released: the papers cite data-use agreements and internal policy.","signals":["ppg"],"subjects":32000,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/samsung-galaxy-watch-pretraining-corpus","links":[{"href":"https://doi.org/10.48550/arXiv.2601.12215","label":"Download","ariaLabel":"Download for Samsung Galaxy Watch PPG pretraining corpus (~32K users, ~17M 10-s segments; private) (opens in new tab)"}]},{"slug":"samsung-mx-ectopic","name":"Samsung MX free-living ectopic-beat set","tagline":"Galaxy Watch PPG (private)","summary":"Free-living watch PPG with ectopic-beat labels. The labeling source is not detailed in xMAE.","signals":["ppg"],"subjects":2613,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/samsung-mx-ectopic","links":[{"href":"https://doi.org/10.48550/arXiv.2605.00973","label":"Download","ariaLabel":"Download for Samsung MX free-living ectopic-beat set (opens in new tab)"}]},{"slug":"scientisst-move-biosignals","name":"ScientISST MOVE: Annotated Wearable Multimodal Biosignals recorded during Everyday Life Activities in Naturalistic Environments","tagline":null,"summary":"ScientISST MOVE records biosignals during everyday activities such as lifting a chair, greeting, gesturing, walking and running. Data came from a chest band, an armband and an Empatica E4 wristband, giving multi-channel electrodermal activity, PPG and ECG. It targets the gap between lab-recorded data and the messier signals seen in daily life.","signals":["ppg","ecg","accel","eda"],"subjects":17,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/scientisst-move-biosignals","links":[{"href":"https://physionet.org/content/scientisst-move-biosignals/1.0.1/","label":"Download","ariaLabel":"Download for ScientISST MOVE: Annotated Wearable Multimodal Biosignals recorded during Everyday Life Activities in Naturalistic Environments (opens in new tab)"},{"href":"https://doi.org/10.13026/hyxq-r919","label":"DOI","ariaLabel":"DOI for ScientISST MOVE: Annotated Wearable Multimodal Biosignals recorded during Everyday Life Activities in Naturalistic Environments (opens in new tab)"}]},{"slug":"sdb","name":"SDB","tagline":"Phone Oximeter pediatric sleep-disordered breathing (Garde 2014)","summary":"This dataset holds phone-oximeter recordings for 146 subjects, with pleth (PPG) at 62.5 Hz and SpO2 at 1 Hz, plus a file of apnea-hypopnea index (AHI) values. It is used for sleep-disordered breathing research.","signals":["ppg","spo2"],"subjects":146,"license":"CC BY 4.0","gated":null,"updated":"2026-10-15","url":"/research/datasets/sdb","links":[{"href":"https://doi.org/10.6084/m9.figshare.1209662","label":"Download","ariaLabel":"Download for SDB (opens in new tab)"}]},{"slug":"senshar","name":"SensHAR","tagline":"wrist-worn IMU (accel/gyro) for six-activity HAR (ESP32-S3 + QMI8658A, n=30)","summary":"Wrist-worn six-axis IMU dataset tied to the SensHAR human-in-the-loop HAR framework (Soumyajit Basak; created 2026-10-14). Custom ESP32-S3 wristband with QMI8658A IMU at 20 Hz; right-wrist accel + gyro; six labeled activities (walk, run, sit, stand, stair down/up). Study n=30 (20 generalized / 10 personalization); archive ships Combined20right.csv + S1right.csv…S10right.csv (~13.85 MB zip).","signals":["accel","gyro"],"subjects":30,"license":"CC BY 4.0","gated":null,"updated":"2026-10-16","url":"/research/datasets/senshar","links":[{"href":"https://ieee-dataport.org/documents/senshar-wrist-worn-inertial-sensor-data-human-activity-recognition","label":"Download","ariaLabel":"Download for SensHAR (opens in new tab)"},{"href":"https://doi.org/10.21227/31y2-7819","label":"DOI","ariaLabel":"DOI for SensHAR (opens in new tab)"}]},{"slug":"sensors-aguirre","name":"Sensors Aguirre 2021","tagline":"PPG→ABP morphology (MIMIC-III processed)","summary":"This dataset pairs arterial blood pressure signals with matching finger PPG. It is a processed version of the MIMIC-III Waveform Database Matched Subset and is used for blood pressure estimation from PPG.","signals":["ppg","abp"],"subjects":1195,"license":"CC BY 4.0","gated":null,"updated":"2026-10-15","url":"/research/datasets/sensors-aguirre","links":[{"href":"https://zenodo.org/records/4598938","label":"Download","ariaLabel":"Download for Sensors Aguirre 2021 (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.4598938","label":"DOI","ariaLabel":"DOI for Sensors Aguirre 2021 (opens in new tab)"}]},{"slug":"senssmarttech","name":"SensSmartTech","tagline":"ECG+PCG+dual-λ PPG+ACC polycardiograph (n=32, 338×30s)","summary":"SensSmartTech (Lazović et al.; PhysioNet v1.0.0, published 2024-12-19). First multisensory base systematically following heart relaxation dynamics across a wide HR range. CSV = original rates; WFDB = resampled to 1 kHz common axis.","signals":["ppg","ecg","accel","pcg"],"subjects":32,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/senssmarttech","links":[{"href":"https://physionet.org/content/senssmarttech/1.0.0/","label":"Download","ariaLabel":"Download for SensSmartTech (opens in new tab)"},{"href":"https://doi.org/10.13026/fy9p-n277","label":"DOI","ariaLabel":"DOI for SensSmartTech (opens in new tab)"}]},{"slug":"shdb-af","name":"SHDB-AF","tagline":"Japanese Holter ECG atrial fibrillation database","summary":"SHDB-AF (Tsutsui / Biton Brimer / Behar et al.; Sci Data 2025 10.1038/s41597-025-04777-4). Built to stress-test AF event detection under distribution shift. Collected May 2019–May 2023 (Saitama).","signals":["ecg"],"subjects":122,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/shdb-af","links":[{"href":"https://physionet.org/content/shdb-af/1.0.1/","label":"Download","ariaLabel":"Download for SHDB-AF (opens in new tab)"},{"href":"https://doi.org/10.13026/n6yq-fq90","label":"DOI","ariaLabel":"DOI for SHDB-AF (opens in new tab)"}]},{"slug":"sicbaldi-wrist-bp-stratification","name":"Sicbaldi wrist PPG BP stratification","tagline":"E4 lab + 24 h free-living vs ABPM (n=39)","summary":"Companion repo to Sicbaldi et al., CinC 2024 (doi 10.22489/CinC.2024.438): Empatica E4 wrist PPG (64 Hz) + ACC (32 Hz) in-lab quiet sitting and 24 h free-living, split into wake and sleep, against Mobil-O-Graph 24 h ABPM; 23 pyPPG morphology features, low vs high BP groups and hour-by-hour trends. Data not hosted in repo → request from authors (gated). 24 h is one day, not multi-day; repeated ABPM readings within that day make it a BP-trend (diurnal) candidate.","signals":["ppg"],"subjects":39,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/sicbaldi-wrist-bp-stratification","links":[{"href":"https://github.com/marcellosicbaldi/BloodPressure_stratification_PPG","label":"Download","ariaLabel":"Download for Sicbaldi wrist PPG BP stratification (opens in new tab)"}]},{"slug":"simultaneous-measurements","name":"Simultaneous physiological measurements with five devices at different cognitive and physical loads","tagline":null,"summary":"This dataset recorded physiology with five devices at the same time: NeXus-10 MKII, eMotion Faros 360, SOMNOtouch NIBP, Hexoskin Hx1 and Polar RS800 Multi. It was made to compare the function, accuracy and usability of ECG devices under different cognitive and physical loads.","signals":["ppg","ecg","accel","resp","spo2","hr"],"subjects":null,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/simultaneous-measurements","links":[{"href":"https://physionet.org/content/simultaneous-measurements/1.0.2/","label":"Download","ariaLabel":"Download for Simultaneous physiological measurements with five devices at different cognitive and physical loads (opens in new tab)"},{"href":"https://doi.org/10.13026/wce5-fj54","label":"DOI","ariaLabel":"DOI for Simultaneous physiological measurements with five devices at different cognitive and physical loads (opens in new tab)"}]},{"slug":"skoda-assembly","name":"Skoda mini checkpoint","tagline":"arm accelerometers, car-assembly gestures (n=1)","summary":"Car-assembly checkpoint gestures with accelerometers on both arms (2008). Source page not re-checked; all facts from Awesome-IMU-Sensing (1 subject).","signals":["accel"],"subjects":null,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/skoda-assembly","links":[{"href":"http://har-dataset.org/","label":"Download","ariaLabel":"Download for Skoda mini checkpoint (opens in new tab)"}]},{"slug":"shhs","name":"Sleep Heart Health Study (SHHS, NSRR)","tagline":null,"summary":"The Sleep Heart Health Study is a multi-center NHLBI cohort that tests whether sleep-disordered breathing raises the risk of heart disease, stroke, hypertension and death. It enrolled 6,441 adults aged 40 or older in 1995 to 1998, and 3,295 had a second polysomnogram in 2001 to 2003. Cardiovascular outcomes were followed until 2011, and the data are shared through the National Sleep Research Resource.","signals":["psg","ecg","eeg","resp","spo2"],"subjects":null,"license":"Not stated","gated":null,"updated":"2026-10-16","url":"/research/datasets/shhs","links":[{"href":"https://sleepdata.org/datasets/shhs","label":"Download","ariaLabel":"Download for Sleep Heart Health Study (SHHS, NSRR) (opens in new tab)"},{"href":"https://doi.org/10.25822/ghy8-ks59","label":"DOI","ariaLabel":"DOI for Sleep Heart Health Study (SHHS, NSRR) (opens in new tab)"}]},{"slug":"sleep-edfx","name":"Sleep-EDF Database Expanded (PhysioNet)","tagline":null,"summary":"Sleep-EDF Expanded contains 197 whole-night polysomnography recordings with EEG, EOG, chin EMG and event markers, and some records also include respiration and body temperature. Trained technicians scored the hypnograms by hand using the Rechtschaffen and Kales manual.","signals":["psg","eeg","resp","temp"],"subjects":null,"license":"Not stated","gated":null,"updated":"2026-10-16","url":"/research/datasets/sleep-edfx","links":[{"href":"https://physionet.org/content/sleep-edfx/1.0.0/","label":"Download","ariaLabel":"Download for Sleep-EDF Database Expanded (PhysioNet) (opens in new tab)"},{"href":"https://doi.org/10.13026/C2X676","label":"DOI","ariaLabel":"DOI for Sleep-EDF Database Expanded (PhysioNet) (opens in new tab)"}]},{"slug":"ppg-cgm-glucose-zenodo","name":"Smartwatch PPG paired with CGM glucose (Zenodo)","tagline":null,"summary":"This dataset pairs continuous smartwatch PPG with glucose values from a continuous glucose monitor (CGM), so models can be trained to estimate glucose from PPG. The authors also report early experiments suggesting reasonable accuracy may be possible, but say more data from more people is needed.","signals":["ppg","cgm"],"subjects":5,"license":"CC BY 4.0","gated":null,"updated":"2026-10-14","url":"/research/datasets/ppg-cgm-glucose-zenodo","links":[{"href":"https://zenodo.org/records/20691830","label":"Download","ariaLabel":"Download for Smartwatch PPG paired with CGM glucose (Zenodo) (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.20691830","label":"DOI","ariaLabel":"DOI for Smartwatch PPG paired with CGM glucose (Zenodo) (opens in new tab)"}]},{"slug":"sonata","name":"SONATA","tagline":"multimodal physio during emotionally-labeled music","summary":"SONATA provides multimodal physiological signals acquired during emotionally-labeled music listening for affective computing / emotion recognition research. Hosted on Zenodo record 22641079 (DOI 10.5281/zenodo.22641079), license CC BY 4.0. Archive: SONATA.zip (~2.4 GB).","signals":["ppg","ecg","skin-temp","eda","resp"],"subjects":50,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/sonata","links":[{"href":"https://zenodo.org/records/22641079","label":"Download","ariaLabel":"Download for SONATA (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.22641079","label":"DOI","ariaLabel":"DOI for SONATA (opens in new tab)"}]},{"slug":"soochow-wristband-hr-menstrual-ovulation","name":"Soochow wristband HR + menstrual/ovulation records (Mendeley v58stpfcnm)","tagline":null,"summary":"Each participant wore a wristband (Huawei or Fitbit) that could detect continuous heart rate for at least two months to record heart rate. Among them, 84 participants recorded their menstrual periods each month with an average age of 23.52 (±2.81), and 28 recorded the day of ovulation by using LH ovulation strips with an average age of 24.00 (±1.83), starting from day 8 of the menstrual cycle until positive results were obtained. The study collected a total of 80 ovulation records and 469 menstrual records. 12 menopausal women and 15 men were recruited, with the average age of 57.08 (±5.44) and 38.80 (±10.82), and the same wristbands were used to collect heart rate data and compare them with women’s data.","signals":["hr"],"subjects":118,"license":"CC BY 4.0","gated":null,"updated":"2026-10-15","url":"/research/datasets/soochow-wristband-hr-menstrual-ovulation","links":[{"href":"https://data.mendeley.com/datasets/v58stpfcnm/1","label":"Download","ariaLabel":"Download for Soochow wristband HR + menstrual/ovulation records (Mendeley v58stpfcnm) (opens in new tab)"},{"href":"https://doi.org/10.17632/v58stpfcnm.1","label":"DOI","ariaLabel":"DOI for Soochow wristband HR + menstrual/ovulation records (Mendeley v58stpfcnm) (opens in new tab)"}]},{"slug":"srf-cognitive-adaptation-eeg-eda-bvp","name":"SRF cognitive adaptation","tagline":"19-ch EEG + EDA + finger BVP + respiration (30 subj, Nexus-32, rest-period autonomic segments)","summary":"Three tasks (pour and drink, simulated driving, shape sorting) under three SRF conditions. Companion papers: TNSRE 2025 (10.1109/TNSRE.2025.3574133), IJRR 2026 (10.1177/02783649261436484), TNSRE 2026 (10.1109/TNSRE.2026.3658402).","signals":["eeg","eda","ppg","resp"],"subjects":30,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/srf-cognitive-adaptation-eeg-eda-bvp","links":[{"href":"https://doi.org/10.6084/m9.figshare.32948735.v2","label":"Download","ariaLabel":"Download for SRF cognitive adaptation (opens in new tab)"}]},{"slug":"incartdb","name":"St Petersburg INCART 12-lead Arrhythmia Database (PhysioNet)","tagline":null,"summary":"The St Petersburg INCART database has 75 annotated 12-lead ECG recordings taken from 32 Holter records. Each record is 30 minutes long, with every lead sampled at 257 Hz.","signals":["ecg"],"subjects":null,"license":"Not stated","gated":null,"updated":"2026-10-16","url":"/research/datasets/incartdb","links":[{"href":"https://physionet.org/content/incartdb/1.0.0/","label":"Download","ariaLabel":"Download for St Petersburg INCART 12-lead Arrhythmia Database (PhysioNet) (opens in new tab)"},{"href":"https://doi.org/10.13026/C2V88N","label":"DOI","ariaLabel":"DOI for St Petersburg INCART 12-lead Arrhythmia Database (PhysioNet) (opens in new tab)"}]},{"slug":"stanford-ppg","name":"Stanford wrist PPG AF (DeepBeat)","tagline":"Torres-Soto & Ashley 2020, Synapse syn21985690","summary":"Torres-Soto J, Ashley EA. Multi-task deep learning for cardiac rhythm detection in wearable devices. npj Digital Medicine 3, 116 (2020).","signals":["ppg"],"subjects":null,"license":"Non-commercial data use agreement (data); GPL-3.0 (code)","gated":null,"updated":"2026-10-16","url":"/research/datasets/stanford-ppg","links":[{"href":"https://www.synapse.org/Synapse:syn21985690","label":"Download","ariaLabel":"Download for Stanford wrist PPG AF (DeepBeat) (opens in new tab)"},{"href":"https://doi.org/10.1038/s41746-020-00320-4","label":"DOI","ariaLabel":"DOI for Stanford wrist PPG AF (DeepBeat) (opens in new tab)"}]},{"slug":"stillness-2025-deichman","name":"Stillness 2025 (Deichman)","tagline":"Equivital group standstill ECG+ACC+resp (n=16)","summary":"Stillness 2025 Deichman (Jensenius / Serdar / Guo / Bjerkestrand / Johnson / Rao; University of Oslo / RITMO). Zenodo concept 21840411; latest version 1.0.4 record 21964815 (2026-08-16). Part of Oslo Standstill Database family.","signals":["ecg","accel","resp","skin-temp","hr","rr-interval"],"subjects":16,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/stillness-2025-deichman","links":[{"href":"https://zenodo.org/records/21964815","label":"Download","ariaLabel":"Download for Stillness 2025 (Deichman) (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.21964815","label":"DOI","ariaLabel":"DOI for Stillness 2025 (Deichman) (opens in new tab)"}]},{"slug":"stress-predict","name":"Stress-Predict Dataset (E4, MIT)","tagline":null,"summary":"Pilot stress-monitoring dataset from Sensors 2022 (10.3390/s22218135). Open GitHub release by Iqbal et al. Three stress-inducing tasks + rest blocks.","signals":["ppg","accel","skin-temp","eda"],"subjects":35,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/stress-predict","links":[{"href":"https://github.com/italha-d/Stress-Predict-Dataset","label":"Download","ariaLabel":"Download for Stress-Predict Dataset (E4, MIT) (opens in new tab)"},{"href":"https://doi.org/10.3390/s22218135","label":"DOI","ariaLabel":"DOI for Stress-Predict Dataset (E4, MIT) (opens in new tab)"}]},{"slug":"stuba-ppg-bgl","name":"STU-BA PPG Blood Glucose Dataset (Zenodo)","tagline":null,"summary":"This open database pairs PPG signals with reference capillary blood glucose readings. Data came from 21 healthy controls and 3 people with type 1 diabetes, recorded with an in-house six-channel ePPG device at 530, 655 and 940 nm. The release also includes record-level signal quality-control files.","signals":["ppg","accel"],"subjects":24,"license":"CC0 1.0","gated":false,"updated":"2026-10-14","url":"/research/datasets/stuba-ppg-bgl","links":[{"href":"https://zenodo.org/records/21978226","label":"Download","ariaLabel":"Download for STU-BA PPG Blood Glucose Dataset (Zenodo) (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.21978226","label":"DOI","ariaLabel":"DOI for STU-BA PPG Blood Glucose Dataset (Zenodo) (opens in new tab)"}]},{"slug":"tempid-fertile-ground-skin-temp","name":"TempID fertile-ground skin temperature + ovulation labels (Zenodo 20160611)","tagline":null,"summary":"This Zenodo deposit accompanies the pilot study “Monitoring female fertile ground: a pilot study of continuous body temperature measurement using a skin wearable device and machine learning integration” by Siret Läänelaid and colleagues at the University of Tartu, the University of Granada and TempID. It contains anonymized, cleaned, filtered, annotated and imputed skin-temperature recordings with ovulation annotations (temperature_ovulation_dataset.csv). A second file (MaxRange.csv) holds hourly mean temperatures computed within the study's maximum temperature range, which served as the input features for its machine-learning pipeline.","signals":["skin-temp"],"subjects":39,"license":"CC BY 4.0","gated":null,"updated":"2026-10-16","url":"/research/datasets/tempid-fertile-ground-skin-temp","links":[{"href":"https://zenodo.org/records/20160611","label":"Download","ariaLabel":"Download for TempID fertile-ground skin temperature + ovulation labels (Zenodo 20160611) (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.20160611","label":"DOI","ariaLabel":"DOI for TempID fertile-ground skin temperature + ovulation labels (Zenodo 20160611) (opens in new tab)"}]},{"slug":"texatrode-textile-holter-ecg-sqa","name":"Texatrode long-term ECG","tagline":"textile vs gel electrodes, 7-day 3-lead Holter + signal-quality labels (13 post-AF-ablation pts; embargoed to 2027-03-31)","summary":"Data for \"Introduction to electrocardiogram signal quality assessment and estimated accuracy for textile electrodes\" (Sci Rep 2025, 10.1038/s41598-025-25365-x). Its SQA sorts segments into 4 quality classes by the estimated accuracy of RR intervals.","signals":["ecg"],"subjects":13,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/texatrode-textile-holter-ecg-sqa","links":[{"href":"https://zenodo.org/records/22874770","label":"Download","ariaLabel":"Download for Texatrode long-term ECG (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.22874770","label":"DOI","ariaLabel":"DOI for Texatrode long-term ECG (opens in new tab)"}]},{"slug":"three-wearable","name":"Three-wearable","tagline":"E4 / Shimmer3 / EmotiBit real-world physio","summary":"This dataset compares wearables in two real-world recording setups. In the first, an Empatica E4, a Shimmer3 GSR+ and an EmotiBit recorded together so the devices can be compared directly. In the second, the Shimmer3 GSR+ and EmotiBit were placed at different body locations to study placement effects.","signals":["ppg","accel","skin-temp","eda"],"subjects":null,"license":"CC BY 4.0","gated":null,"updated":"2026-10-14","url":"/research/datasets/three-wearable","links":[{"href":"https://zenodo.org/records/19218368","label":"Download","ariaLabel":"Download for Three-wearable (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.19218368","label":"DOI","ariaLabel":"DOI for Three-wearable (opens in new tab)"}]},{"slug":"tnda-har","name":"TNDA-HAR","tagline":"wrist/ankle/back IMU, 8 activities (n=23)","summary":"IEEE DataPort open access (doi:10.21227/4epb-pg26). Instructions: 23 subjects, 8 activities (sit, stand, lay, walk, run, cycle, stairs up/down), ~2 min each; acc/gyro/mag at wrist, ankle and back. CSV, 353.7 MB zip.","signals":["accel"],"subjects":23,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/tnda-har","links":[{"href":"https://ieee-dataport.org/open-access/tnda-har-0","label":"Download","ariaLabel":"Download for TNDA-HAR (opens in new tab)"}]},{"slug":"tolife-gw5","name":"ToLife","tagline":"paired Galaxy Watch 5 wrist PPG, finger PPG, and ECG","summary":"ToLife provides simultaneous consumer-watch wrist PPG, contact finger PPG, and ECG. It is especially useful for learning timing/rate relationships and testing cross-site representations without assuming that fingertip contour or amplitude is preserved at the wrist.","signals":["ppg","ecg"],"subjects":20,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/tolife-gw5","links":[{"href":"https://zenodo.org/records/10717903","label":"Download","ariaLabel":"Download for ToLife (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.10717903","label":"DOI","ariaLabel":"DOI for ToLife (opens in new tab)"}]},{"slug":"tolife-csa-health-indices","name":"TOLIFE Clinical Study A","tagline":"sensor-derived health indices (COPD, 12-month)","summary":"This dataset comes from Clinical Study A of the EU TOLIFE project on COPD, which used AI and smart sensing to improve care. It contains sensor-derived health indices for participants monitored at sites in Germany, Spain and Italy from May 2024 to April 2026.","signals":["derived-health-indices"],"subjects":86,"license":"CC BY 4.0","gated":null,"updated":"2026-10-14","url":"/research/datasets/tolife-csa-health-indices","links":[{"href":"https://zenodo.org/records/23263785","label":"Download","ariaLabel":"Download for TOLIFE Clinical Study A (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.23263785","label":"DOI","ariaLabel":"DOI for TOLIFE Clinical Study A (opens in new tab)"}]},{"slug":"shimmer-dehydration-sabry","name":"Towards On-Device Dehydration Monitoring Using Machine Learning from Wearable Device's Data (Shimmer3 GSR+ raw dataset)","tagline":null,"summary":"This dataset was recorded with a Shimmer3 GSR+ unit on 11 subjects for on-device dehydration monitoring. It includes accelerometer, gyroscope, magnetometer, skin conductance, PPG, temperature and barometric pressure. Some recordings were made during Muslim fasting of up to 15 hours without food or water, with the time since last drink logged for each sample.","signals":["ppg","eda","accel","gyro","mag","skin-temp","baro"],"subjects":11,"license":"CC BY 4.0","gated":null,"updated":"2026-10-15","url":"/research/datasets/shimmer-dehydration-sabry","links":[{"href":"https://doi.org/10.5281/zenodo.7099266","label":"Download","ariaLabel":"Download for Towards On-Device Dehydration Monitoring Using Machine Learning from Wearable Device's Data (Shimmer3 GSR+ raw dataset) (opens in new tab)"}]},{"slug":"alfaro-treadmill-fatigue-polar-ppg-eda","name":"Treadmill fatigue","tagline":"Polar H10 HR/RR + secondary PPG + GSR with Borg CR10 (Alfaro et al.)","summary":"This dataset records wearable signals, treadmill data and perceived exertion from six healthy adults during two lab protocols that build up physical fatigue. It was made to develop estimators of perceived exertion, exertion state and time to exhaustion, for assigning tasks in human-robot collaboration. Signals include Polar chest-strap heart rate and RR intervals, PPG heart rate and inter-beat intervals with quality flags, and skin conductance, all raw and unfiltered with artefacts left in.","signals":["ecg","ppg","eda","hr"],"subjects":6,"license":"CC BY 4.0","gated":null,"updated":"2026-10-14","url":"/research/datasets/alfaro-treadmill-fatigue-polar-ppg-eda","links":[{"href":"https://zenodo.org/records/23208045","label":"Download","ariaLabel":"Download for Treadmill fatigue (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.23208045","label":"DOI","ariaLabel":"DOI for Treadmill fatigue (opens in new tab)"}]},{"slug":"triple-wavelength-hgb-abuzairi","name":"Triple-Wavelength PPG for Hemoglobin (Abuzairi / Mendeley)","tagline":null,"summary":"This dataset contains triple-wavelength PPG recordings for estimating hemoglobin. Data were collected from 52 subjects (60% female) in Depok City, West Java, Indonesia.","signals":["ppg"],"subjects":52,"license":"CC BY 4.0","gated":null,"updated":"2026-10-14","url":"/research/datasets/triple-wavelength-hgb-abuzairi","links":[{"href":"https://data.mendeley.com/datasets/8xbbck45tt/2","label":"Download","ariaLabel":"Download for Triple-Wavelength PPG for Hemoglobin (Abuzairi / Mendeley) (opens in new tab)"},{"href":"https://doi.org/10.17632/8xbbck45tt.2","label":"DOI","ariaLabel":"DOI for Triple-Wavelength PPG for Hemoglobin (Abuzairi / Mendeley) (opens in new tab)"}]},{"slug":"tutunji-prolonged-stress-e4","name":"Tutunji prolonged stress","tagline":"Empatica E4 + EMA over exam vs control weeks (n=83, Radboud)","summary":"Collection Detecting Prolonged Stress in Real Life Using Wearable Biosensors and Ecological Momentary Assessments: Naturalistic Experimental Study (version 1), Radboud University / Donders Centre for Cognitive Neuroimaging, collection di.dccn.DSC3013068.02912. Authors: Tutunji R., Kogias N., Kapteijns B., Krentz M., Krause F., Vassena E., Hermans E.J. Companion paper DOI 10.2196/39995 (JMIR).","signals":["ppg","accel","eda","skin-temp","hr","ibi"],"subjects":83,"license":"Not stated","gated":null,"updated":"2026-10-16","url":"/research/datasets/tutunji-prolonged-stress-e4","links":[{"href":"https://data.ru.nl/collections/di/dccn/DSC_3013068.02_912","label":"Download","ariaLabel":"Download for Tutunji prolonged stress (opens in new tab)"},{"href":"https://doi.org/10.34973/j0ym-mb28","label":"DOI","ariaLabel":"DOI for Tutunji prolonged stress (opens in new tab)"}]},{"slug":"two-wearables-placements","name":"Two-wearables placements","tagline":"EmotiBit + Shimmer3 wrist/finger","summary":"Placement-focused Scenario 2 release of a multi-wearable comparison study (Empatica E4 simultaneous Scenario 1 not included in this version). Zenodo 22655618, DOI 10.5281/zenodo.22655618, license CC BY 4.0. Companion processing: https://github.com/gadea-lucas/wearables-signal-comparison","signals":["ppg","accel","gyro","magnetometer","skin-temp","eda"],"subjects":10,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/two-wearables-placements","links":[{"href":"https://zenodo.org/records/22655618","label":"Download","ariaLabel":"Download for Two-wearables placements (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.22655618","label":"DOI","ariaLabel":"DOI for Two-wearables placements (opens in new tab)"}]},{"slug":"ubfc-phys","name":"UBFC-Phys","tagline":"multimodal social-stress (wrist BVP+EDA + face video)","summary":"UBFC-Phys (Université Bourgogne Franche-Comté / Meziati Sabour et al.). Companion paper in IEEE Transactions on Affective Computing 2021.","signals":["eda","ppg-bvp","video-face"],"subjects":56,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/ubfc-phys","links":[{"href":"https://sites.google.com/view/ybenezeth/ubfc-phys","label":"Download","ariaLabel":"Download for UBFC-Phys (opens in new tab)"}]},{"slug":"uci-cuffless-bp","name":"UCI Cuff-Less Blood Pressure Estimation (UCI 340; Kachuee 2015)","tagline":null,"summary":"Kachuee, Kiani, Mohammadzade, Shabany. Introductory paper: \"Cuff-less high-accuracy calibration-free blood pressure estimation using pulse transit time\" (ISCAS 2015). The UCI page says the raw ECG, PPG and ABP were \"originally collected from the physionet.org\" and then preprocessed and validated.","signals":["ppg","abp","ecg"],"subjects":null,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/uci-cuffless-bp","links":[{"href":"https://archive.ics.uci.edu/dataset/340/cuff+less+blood+pressure+estimation","label":"Download","ariaLabel":"Download for UCI Cuff-Less Blood Pressure Estimation (UCI 340; Kachuee 2015) (opens in new tab)"},{"href":"https://doi.org/10.24432/C5B602","label":"DOI","ariaLabel":"DOI for UCI Cuff-Less Blood Pressure Estimation (UCI 340; Kachuee 2015) (opens in new tab)"}]},{"slug":"uci-har","name":"UCI Human Activity Recognition Using Smartphones (UCI-HAR)","tagline":null,"summary":"UCI-HAR records 30 volunteers aged 19 to 48 doing six daily activities (walking, walking upstairs, walking downstairs, sitting, standing and lying) with a Samsung Galaxy S II worn on the waist. The phone's accelerometer and gyroscope captured 3-axis acceleration and angular velocity at 50 Hz, and labels were added manually from video. The data are split 70/30 by volunteer into training and test sets.","signals":["imu"],"subjects":30,"license":"Creative Commons Attribution 4.0 International (CC BY 4.0)","gated":false,"updated":"2026-10-16","url":"/research/datasets/uci-har","links":[{"href":"https://archive.ics.uci.edu/dataset/240/human+activity+recognition+using+smartphones","label":"Download","ariaLabel":"Download for UCI Human Activity Recognition Using Smartphones (UCI-HAR) (opens in new tab)"},{"href":"https://doi.org/10.24432/C54S4K","label":"DOI","ariaLabel":"DOI for UCI Human Activity Recognition Using Smartphones (UCI-HAR) (opens in new tab)"}]},{"slug":"ucsf-icu-waveforms-private","name":"UCSF ICU bedside-monitor waveforms (private; GPT-PPG / SiamQuality / SQUWA pretraining)","tagline":null,"summary":"Continuous bedside-monitor waveforms plus arrhythmia alarms and EHR from UCSF Medical Center ICUs, collected March 2013 – December 2018. Used under UCSF IRB 14-13262 (waiver of consent). The Emory authors access it through a UCSF–Emory data-use agreement.","signals":["ppg","ecg","resp"],"subjects":24100,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/ucsf-icu-waveforms-private","links":[{"href":"https://doi.org/10.48550/arXiv.2503.08015","label":"Download","ariaLabel":"Download for UCSF ICU bedside-monitor waveforms (private; GPT-PPG / SiamQuality / SQUWA pretraining) (opens in new tab)"}]},{"slug":"uk-biobank-accelerometer","name":"UK Biobank wrist accelerometer (Axivity AX3 / field 90001)","tagline":null,"summary":"UK Biobank category 1008 / field 90001 “Acceleration data – cwa format”. Main collection ~2013–2016; seasonal repeats for a subset. Official field page: https://biobank.ctsu.ox.ac.uk/ukb/field.cgi?id=90001.","signals":["accel"],"subjects":103568,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/uk-biobank-accelerometer","links":[{"href":"https://biobank.ctsu.ox.ac.uk/ukb/field.cgi?id=90001","label":"Download","ariaLabel":"Download for UK Biobank wrist accelerometer (Axivity AX3 / field 90001) (opens in new tab)"}]},{"slug":"umafall","name":"UMAFall","tagline":"multi-sensor ADL + fall traces (Universidad de Málaga, n=19)","summary":"Mobility traces from 19 experimental subjects emulating predetermined ADLs and falls; smartphone plus wearable motion sensors; video clips of movements included. Must cite Casilari et al. 2017 and Santoyo-Ramón et al.","signals":["accel"],"subjects":19,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/umafall","links":[{"href":"https://figshare.com/articles/dataset/UMA_ADL_FALL_Dataset_zip/4214283","label":"Download","ariaLabel":"Download for UMAFall (opens in new tab)"}]},{"slug":"universe-mwl-stress","name":"UNIVERSE (mental workload + stress, ~315h E4)","tagline":null,"summary":"UNIVERSE was recorded from 24 participants during an eight-hour cognitive load paradigm using consumer-grade wearables. The anonymised data and extracted features are public and suit work on real-time workload detection and artifact removal from low-cost devices.","signals":["ppg","accel","skin-temp","eda","eeg"],"subjects":24,"license":"CC BY 4.0","gated":null,"updated":"2026-10-14","url":"/research/datasets/universe-mwl-stress","links":[{"href":"https://zenodo.org/records/10371068","label":"Download","ariaLabel":"Download for UNIVERSE (mental workload + stress, ~315h E4) (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.10371068","label":"DOI","ariaLabel":"DOI for UNIVERSE (mental workload + stress, ~315h E4) (opens in new tab)"}]},{"slug":"uqvitalsigns","name":"UQVitalSigns","tagline":"University of Queensland anesthesia vital signs","summary":"32 surgical cases; multi-parameter OR monitoring including pulse-ox pleth. CC BY-NC 3.0 → stored under raw/restricted/. Not for pretrain allowlist.","signals":["ppg","ecg","resp","spo2","co2","bp"],"subjects":32,"license":"CC BY-NC 3.0","gated":null,"updated":"2026-10-15","url":"/research/datasets/uqvitalsigns","links":[{"href":"https://outbox.eait.uq.edu.au/uqdliu3/uqvitalsignsdataset/","label":"Download","ariaLabel":"Download for UQVitalSigns (opens in new tab)"},{"href":"https://doi.org/10.1213/ANE.0b013e318241f7c0","label":"DOI","ariaLabel":"DOI for UQVitalSigns (opens in new tab)"}]},{"slug":"usc-had","name":"USC-HAD","tagline":"hip MotionNode acc/gyro/mag daily activities (n=14)","summary":"USC-SIPI Human Activity Dataset (UbiComp SAGAware 2012). Page states no licence. MotionNode IMU at hip; N=14 per Awesome-IMU-Sensing (paper not re-read).","signals":["accel"],"subjects":14,"license":"Not stated","gated":false,"updated":"2026-10-14","url":"/research/datasets/usc-had","links":[{"href":"https://sipi.usc.edu/had/","label":"Download","ariaLabel":"Download for USC-HAD (opens in new tab)"}]},{"slug":"utsa-ppg","name":"UTSA-PPG","tagline":"multimodal wrist+finger PPG HR/HRV (Xu 2025)","summary":"UTSA-PPG is a multimodal dataset for heart rate and HRV research. Twelve subjects each did three sessions of at least 30 minutes (sitting, sleeping and office work) while wearing custom finger PPG (infrared and red) and wrist PPG (green) sensors with an accelerometer, a Fitbit Charge 6, an O2Ring oximeter and a 3-lead ECG Holter as reference. Fifteen people took part, but three were dropped after ECG device failures.","signals":["ppg","ecg","accel"],"subjects":12,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/utsa-ppg","links":[{"href":"https://github.com/utsanpb2-234/UTSA-PPG","label":"Download","ariaLabel":"Download for UTSA-PPG (opens in new tab)"},{"href":"https://doi.org/10.1109/ICHI64645.2025.00042","label":"DOI","ariaLabel":"DOI for UTSA-PPG (opens in new tab)"}]},{"slug":"vitaldb","name":"VitalDB","tagline":"intraoperative multi-parameter vitals (PhysioNet mirror)","summary":null,"signals":["ppg","ecg","resp","abp","co2"],"subjects":6388,"license":"CC BY 4.0 (PhysioNet release)","gated":false,"updated":"2026-10-15","url":"/research/datasets/vitaldb","links":[{"href":"https://physionet.org/content/vitaldb/1.0.0/","label":"Download","ariaLabel":"Download for VitalDB (opens in new tab)"},{"href":"https://doi.org/10.13026/czw8-9p62","label":"DOI","ariaLabel":"DOI for VitalDB (opens in new tab)"}]},{"slug":"vitaldb-arrhythmia","name":"VitalDB Arrhythmia","tagline":"anesthesiologist-validated intraoperative ECG beat/rhythm labels (n=482)","summary":"The VitalDB Arrhythmia Database is a set of annotated ECG recordings taken during surgery, for building and testing arrhythmia detectors in the operating room. A custom deep learning beat classifier screened the source data for candidate segments, and each segment was then reviewed by at least two of five anesthesiologists.","signals":["ecg"],"subjects":482,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/vitaldb-arrhythmia","links":[{"href":"https://physionet.org/content/vitaldb-arrhythmia/1.0.0/","label":"Download","ariaLabel":"Download for VitalDB Arrhythmia (opens in new tab)"},{"href":"https://doi.org/10.13026/axd6-wm13","label":"DOI","ariaLabel":"DOI for VitalDB Arrhythmia (opens in new tab)"}]},{"slug":"vitalvideos","name":"VitalVideos","tagline":"face video + contact PPG + BP/SpO2 (Europe-1, Africa-1, Worldwide)","summary":"VitalVideos provides facial video datasets with synchronized physiological ground truth for remote photoplethysmography (rPPG) research. The original Europe-1 set has 850 participants with 30-second, 30 fps raw video and PPG, heart rate, SpO2 and blood pressure references, and an Africa-1 set adds 1,550 participants. Contact PPG comes from a CMS50D+ oximeter at about 60 Hz.","signals":["ppg"],"subjects":850,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/vitalvideos","links":[{"href":"https://vitalvideos.org/","label":"Download","ariaLabel":"Download for VitalVideos (opens in new tab)"}]},{"slug":"w-har","name":"w-HAR","tagline":"ankle IMU + knee stretch sensor HAR","summary":"w-HAR is a human activity recognition dataset and framework built around low-power wearable devices, published by Bhat, Tran, Shill and Ogras in Sensors in 2020. It pairs an ankle-worn IMU with a stretch sensor at the knee, and the GitHub repository provides the data together with source code for online activity recognition.","signals":["accel"],"subjects":null,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/w-har","links":[{"href":"https://github.com/gmbhat/human-activity-recognition","label":"Download","ariaLabel":"Download for w-HAR (opens in new tab)"}]},{"slug":"walker-posture-e4","name":"Walker posture E4","tagline":"Empatica E4 + MoCap + smart rollator (Zenodo 22734705)","summary":"“Evaluating the validity of walker sensors for posture monitoring” (Fernández-Carmona / Urdiales / Gómez-de-Gabriel; Universidad de Málaga / TaISLab). ROS2 bags converted to per-trial CSV zips (202404test.zip, 202406test.zip). Code: https://github.com/TaISLab/WalKit.","signals":["ppg","accel","skin-temp","eda"],"subjects":null,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/walker-posture-e4","links":[{"href":"https://zenodo.org/records/22734705","label":"Download","ariaLabel":"Download for Walker posture E4 (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.22734705","label":"DOI","ariaLabel":"DOI for Walker posture E4 (opens in new tab)"}]},{"slug":"warehouse-moc-sensor","name":"Warehouse MoC+Sensor","tagline":"full-body MoCap + real wrist Garmin accelerometer","summary":"Rokoko full-body MoCap suit time-aligned with a physical right-wrist Garmin Venu accelerometer; warehouse/logistics actions manually segmented with synchronized video and beat instruction. Repo licence MIT; data-specific terms not separately stated — verdict left unknown pending confirmation. Awesome list: 9 subjects, 4.4 h.","signals":["accel"],"subjects":9,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/warehouse-moc-sensor","links":[{"href":"https://github.com/xieyulai/warehouse-moc-sensor","label":"Download","ariaLabel":"Download for Warehouse MoC+Sensor (opens in new tab)"}]},{"slug":"warm-vr","name":"WARM-VR","tagline":"Empatica E4 BVP/EDA/TEMP/ACC + Polar H10 ECG during VR stress and multisensory relaxation (n=28)","summary":"Alghoul, Mohd (Faisal), Laamarti, Al Osman, El Saddik (University of Ottawa). ACM TOMM 2026, DOI 10.1145/3836767; arXiv 2605.00184. Dataset DOI 10.21227/rwrm-n531.","signals":["ppg","eda","skin-temp","accel","ecg"],"subjects":28,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/warm-vr","links":[{"href":"https://doi.org/10.21227/rwrm-n531","label":"Download","ariaLabel":"Download for WARM-VR (opens in new tab)"}]},{"slug":"watchmyheart","name":"WatchMyHeart","tagline":"Apple Watch 7 + Galaxy Watch 4 resting ECG (n=93)","summary":"WatchMyHeart (Alkhodari / Khandoker, Khalifa University). Zenodo 20729337 (concept 20729336), published 2026-06-22; files modified 2026-08-11. Release: Rawsignals.mat, Metadata.xlsx, Qualitysummaryperparticipant.xlsx.","signals":["ecg"],"subjects":93,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/watchmyheart","links":[{"href":"https://zenodo.org/records/20729337","label":"Download","ariaLabel":"Download for WatchMyHeart (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.20729337","label":"DOI","ariaLabel":"DOI for WatchMyHeart (opens in new tab)"}]},{"slug":"wear-sports","name":"WEAR","tagline":"outdoor sports, wrist + ankle accelerometers + egocentric video (n=22)","summary":"22 participants, 18 workout activities, synchronized acceleration and egocentric video at 11 outdoor locations (IMWUT 8(4) art. 175, 2024). Licence CC BY-NC-SA 4.0 — non-commercial.","signals":["accel"],"subjects":22,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/wear-sports","links":[{"href":"https://mariusbock.github.io/wear/","label":"Download","ariaLabel":"Download for WEAR (opens in new tab)"}]},{"slug":"wearanize-plus-oa","name":"Wearanize+ OA","tagline":"overnight E4 + Zmax + ActivPAL + PSG sleep (n=130)","summary":"Wearanize+ OA is the open-access release of Wearanize+ (Donders / Radboud; Dresler / Zerr / Sikder). Each participant slept one night with three wearables plus PSG. PSG stages include two expert manual scorers plus U-Sleep v2.0 autoscores.","signals":["ppg","accel","skin-temp","eda","eeg","eog","psg"],"subjects":130,"license":"Not stated","gated":null,"updated":"2026-10-16","url":"/research/datasets/wearanize-plus-oa","links":[{"href":"https://data.ru.nl/collections/di/dcmn/DSC_wrnzpoa_t0000925a_195","label":"Download","ariaLabel":"Download for Wearanize+ OA (opens in new tab)"},{"href":"https://doi.org/10.34973/xrmf-5726","label":"DOI","ariaLabel":"DOI for Wearanize+ OA (opens in new tab)"}]},{"slug":"wenxinwuyang-handheld-ecg-ppg","name":"WenXinWuYang","tagline":"HeartVoice handheld single-lead ECG + finger PPG recordings (private; AnyPPG eval)","summary":"Synchronized PPG and single-lead ECG from a handheld consumer device; labels come from the device's ECG algorithms. AnyPPG (arXiv 2511.01747, supplement) describes it. The device itself is validated in Wang H.","signals":["ppg","ecg"],"subjects":null,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/wenxinwuyang-handheld-ecg-ppg","links":[{"href":"https://doi.org/10.48550/arXiv.2511.01747","label":"Download","ariaLabel":"Download for WenXinWuYang (opens in new tab)"}]},{"slug":"wesad","name":"WESAD","tagline":"Wearable Stress and Affect Detection","summary":"WESAD (Schmidt / Reiss / Duerichen / Marberger / Van Laerhoven), ICMI 2018. Lab study with baseline / stress / amusement states plus questionnaires.","signals":["ppg","ecg","accel","skin-temp","eda","resp"],"subjects":15,"license":"Scientific, non-commercial use with credit to the owners","gated":null,"updated":"2026-10-16","url":"/research/datasets/wesad","links":[{"href":"https://ubi29.informatik.uni-siegen.de/usi/data_wesad.html","label":"Download","ariaLabel":"Download for WESAD (opens in new tab)"}]},{"slug":"wf-ppg","name":"WF-PPG","tagline":"wrist + finger multi-wavelength PPG under graded contact pressure (cuff BP per session)","summary":"WF-PPG is a dual-channel dataset that records PPG at the wrist and finger at the same time. It was built to study how wrist posture changes PPG waveform shape.","signals":["ppg","ecg","spo2","contact-pressure"],"subjects":27,"license":"CC BY 4.0","gated":null,"updated":"2026-10-14","url":"/research/datasets/wf-ppg","links":[{"href":"https://doi.org/10.6084/m9.figshare.27011998","label":"Download","ariaLabel":"Download for WF-PPG (opens in new tab)"}]},{"slug":"whamap","name":"WHAMAP","tagline":"multi-IMU lower-limb + phone HAR with Borg fatigue (n=10)","summary":"WHAMAP contains data from five IMUs on the lower limbs (thighs, shanks and right foot) plus smartphone sensors, recorded from 10 participants. They sat, stood, walked, walked uphill and squatted, and the walking and squatting trials include fatigue levels rated on the Borg scale.","signals":["accel","gyro"],"subjects":10,"license":"CC BY 4.0","gated":null,"updated":"2026-10-15","url":"/research/datasets/whamap","links":[{"href":"https://data.mendeley.com/datasets/8zt9g9g64c/1","label":"Download","ariaLabel":"Download for WHAMAP (opens in new tab)"},{"href":"https://doi.org/10.17632/8zt9g9g64c.1","label":"DOI","ariaLabel":"DOI for WHAMAP (opens in new tab)"}]},{"slug":"wharf","name":"WHARF","tagline":"single wrist accelerometer ADL primitives (n=16)","summary":"WHARF (Wearable Human Activity Recognition Folder) is a public collection of labelled recordings from a single wrist-worn tri-axial accelerometer, intended for building and validating models of simple daily activities. It ships with MATLAB functions for creating and testing those models and is described in papers at IEEE ICRA 2013 and IEEE CASE 2012.","signals":["accel"],"subjects":null,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/wharf","links":[{"href":"https://github.com/fulviomas/WHARF","label":"Download","ariaLabel":"Download for WHARF (opens in new tab)"}]},{"slug":"wildppg","name":"WildPPG","tagline":null,"summary":"ETH Zürich (Siplab) real-world PPG dataset — NeurIPS 2024 Datasets & Benchmarks. Four custom devices worn simultaneously at head, sternum, wrist, ankle during a full-day outdoor trip from Zurich to Jungfraujoch (up to 3,571 m elevation).","signals":["ppg","ecg","accel","skin-temp","baro"],"subjects":16,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/wildppg","links":[{"href":"https://siplab.org/projects/WildPPG","label":"Download","ariaLabel":"Download for WildPPG (opens in new tab)"}]},{"slug":"wimid","name":"WIMID","tagline":"Wrist IMU Meals and Intakes","summary":"WIMID (Wrist IMU Meals and Intakes Dataset) from Jožef Stefan Institute (Jordan / Luštrek). Zenodo 10.5281/zenodo.18176217. Preprocessed IMU (interpolated to 100 Hz, median-filtered).","signals":["accel","gyro"],"subjects":11,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/wimid","links":[{"href":"https://zenodo.org/records/18176217","label":"Download","ariaLabel":"Download for WIMID (opens in new tab)"},{"href":"https://doi.org/10.5281/zenodo.18176217","label":"DOI","ariaLabel":"DOI for WIMID (opens in new tab)"}]},{"slug":"wisdm","name":"WISDM AR v1.1","tagline":"cellphone accelerometer HAR (n=36)","summary":"Wireless Sensor Data Mining (WISDM) Lab, Fordham University. Activity Prediction / AR dataset v1.1 corresponds to Kwapisz et al., Activity Recognition using Cell Phone Accelerometers (SensorKDD 2010). Fordham dataset page hosts raw + transformed files; user IDs masked 1–36.","signals":["accel"],"subjects":36,"license":"Not stated","gated":false,"updated":"2026-10-16","url":"/research/datasets/wisdm","links":[{"href":"https://www.cis.fordham.edu/wisdm/dataset.php","label":"Download","ariaLabel":"Download for WISDM AR v1.1 (opens in new tab)"}]},{"slug":"wrist-ppg-exercise","name":"Wrist PPG During Exercise (PhysioNet wrist)","tagline":null,"summary":"PhysioNet wrist PPG during exercise corpus (wrist 1.0.0). Treadmill walk/jog and bike resistance protocols with synchronized chest ECG R-peaks.","signals":["ppg","ecg","accel","gyro"],"subjects":8,"license":"ODC-By 1.0","gated":null,"updated":"2026-10-14","url":"/research/datasets/wrist-ppg-exercise","links":[{"href":"https://physionet.org/content/wrist/1.0.0/","label":"Download","ariaLabel":"Download for Wrist PPG During Exercise (PhysioNet wrist) (opens in new tab)"},{"href":"https://doi.org/10.13026/C2PQ1X","label":"DOI","ariaLabel":"DOI for Wrist PPG During Exercise (PhysioNet wrist) (opens in new tab)"}]},{"slug":"yoga4all-surya-namaskar-wrist-imu","name":"Yoga4All","tagline":"wrist IMU (accel/gyro/mag) for 12 Surya Namaskar postures (21 subj, 107 sessions, 1,098 rounds)","summary":"Yoga4All records wrist motion while people performed the 12 postures of Surya Namaskar, using Arduino Nano 33 BLE Sense boards and ESP32-C6 smartwatches worn mostly on the right wrist. Twenty-one participants contributed 107 sessions and 1,098 rounds over 35 collection days between July and September 2026.","signals":["accel","gyro","magnetometer"],"subjects":21,"license":"CC BY 4.0","gated":null,"updated":"2026-10-14","url":"/research/datasets/yoga4all-surya-namaskar-wrist-imu","links":[{"href":"https://ieee-dataport.org/documents/yoga4all-end-end-framework-collecting-wearable-sensor-datasets-recognize-and-monitor","label":"Download","ariaLabel":"Download for Yoga4All (opens in new tab)"},{"href":"https://doi.org/10.21227/prkc-h546","label":"DOI","ariaLabel":"DOI for Yoga4All (opens in new tab)"}]},{"slug":"yunnan-dual-wavelength-ppg-ecg-pressure-pulse-wave","name":"Yunnan dual-wavelength PPG + ECG + pressure pulse wave (102 subj, 1000 Hz)","tagline":null,"summary":"This open dataset contains synchronized dual-wavelength PPG, ECG and pressure pulse wave signals from 102 participants aged 17 to 93 (53% male). Age, gender, blood pressure, heart rate and partial medical history were also recorded. A heart-rate consistency check gave 86.63% accuracy and 100% precision as a quality measure.","signals":["ppg","ecg","pressure"],"subjects":102,"license":"CC BY 4.0","gated":null,"updated":"2026-10-14","url":"/research/datasets/yunnan-dual-wavelength-ppg-ecg-pressure-pulse-wave","links":[{"href":"https://figshare.com/articles/dataset/_b_Dataset_for_Synchronous_Dual_wavelength_Photoplethysmography_Electrocardiography_and_Pulse_Wave_b_/33033332/1","label":"Download","ariaLabel":"Download for Yunnan dual-wavelength PPG + ECG + pressure pulse wave (102 subj, 1000 Hz) (opens in new tab)"},{"href":"https://doi.org/10.6084/m9.figshare.33033332.v1","label":"DOI","ariaLabel":"DOI for Yunnan dual-wavelength PPG + ECG + pressure pulse wave (102 subj, 1000 Hz) (opens in new tab)"}]},{"slug":"zchsound","name":"ZCHSound","tagline":"ZJU paediatric heart-sound database (congenital heart disease)","summary":"ZCHSound is a pediatric heart sound database for detecting congenital heart disease. Between 2020 and 2022 the authors worked with cardiac surgeons at three children's hospitals to record 1,259 participants with electronic stethoscopes, and two cardiac experts confirmed every label. A random forest ensemble reached an F1 score of 90.3% for normal versus pathological heart sounds.","signals":["audio"],"subjects":1259,"license":"CC BY 4.0","gated":null,"updated":"2026-10-14","url":"/research/datasets/zchsound","links":[{"href":"http://zchsound.ncrcch.org.cn/","label":"Download","ariaLabel":"Download for ZCHSound (opens in new tab)"},{"href":"https://doi.org/10.1109/TBME.2023.3348800","label":"DOI","ariaLabel":"DOI for ZCHSound (opens in new tab)"}]},{"slug":"zu-pwd-25","name":"ZU-PWD '25","tagline":"Embrace Plus free-living wellness (minute-level)","summary":"ZU-PWD ’25 (DOI 10.21227/2tth-wj12) releases synchronized minute-level physiology/behavior from Empatica Embrace Plus with a companion self-report set (ZU-PWD-SR ’25). Authors: Evangelista / Nazir / Bukhari / Dahmani / Tbaishat / Sharma (Zayed University et al.). Created 2025-09-16 on IEEE DataPort; last updated 2025-09-24.","signals":["accel","skin-temp","eda","resp","hr","hrv"],"subjects":28,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/zu-pwd-25","links":[{"href":"https://ieee-dataport.org/documents/zayed-university-physiological-wellness-dataset-2025-zu-pwd-25","label":"Download","ariaLabel":"Download for ZU-PWD '25 (opens in new tab)"},{"href":"https://doi.org/10.21227/2tth-wj12","label":"DOI","ariaLabel":"DOI for ZU-PWD '25 (opens in new tab)"}]},{"slug":"udpad-wrist-ppg-imu-microgesture","name":"μDPad","tagline":"Large-Scale Multimodal PPG and IMU Dataset for Wrist-Worn Microgesture Recognition","summary":"The authors built a dataset because commercial watches with similar multi-photodiode optical layouts don't expose raw channels. The research question is whether spatial PPG picks up the small tissue/tendon deformations under the watch during finger microgestures.","signals":["ppg","accel","gyro"],"subjects":65,"license":"Not stated","gated":null,"updated":"2026-10-15","url":"/research/datasets/udpad-wrist-ppg-imu-microgesture","links":[{"href":"https://huggingface.co/datasets/eth-siplab/microdpad","label":"Download","ariaLabel":"Download for μDPad (opens in new tab)"},{"href":"https://doi.org/10.1145/3776574.3831197","label":"DOI","ariaLabel":"DOI for μDPad (opens in new tab)"}]},{"slug":"tau-ring-dataset","name":"τ-Ring dataset (thuhci/RingTool)","tagline":null,"summary":"The τ-Ring (HealthRing) dataset, published in Scientific Data, records smart-ring PPG and IMU data alongside medical reference devices. Two custom rings with reflective and transmissive optical paths were used in two protocols: resting, simulated low oxygen and squats, and a daily-activity protocol with talking, head movement, standing and walking in place. The RingTool toolkit estimates heart rate, respiratory rate, SpO2 and blood pressure with physics-based and deep learning models.","signals":["accel","ppg-ir","ppg-red"],"subjects":null,"license":"Not stated","gated":null,"updated":"2026-10-14","url":"/research/datasets/tau-ring-dataset","links":[{"href":"https://github.com/thuhci/RingTool","label":"Download","ariaLabel":"Download for τ-Ring dataset (thuhci/RingTool) (opens in new tab)"}]}]}