PhysDrive is a 48-driver multimodal remote physiological measurement dataset for in-vehicle monitoring with synchronized RGB, NIR, and mmWave plus ECG/BVP/RESP/HR/RR/SpO2 (NeurIPS 2025; arXiv:2507.19172). Preprocessed Kaggle slice needs no DSA; full raw RGB/NIR requires academic data-share agreement and commercial use is banned.
- License
Academic-only; commercial use banned; code MIT; raw data under DSA
- Type
- Dataset
- DOI
- Not reported
- Licence
- Academic-only; commercial use banned; code MIT; raw data under DSA
- Access
- Not reported
- Subjects
- 48
- PPG sites
- Not reported
- Setting
- Not reported
- Wrist
- Not reported
- Raw waveforms
- Not reported
- Tasks
- Not reported
- Updated
- Oct 24, 2026
Original description
Here is [NIPS2025] PhysDrive: A Multimodal Remote Physiological Measurement Dataset for In-vehicle Driver Monitoring, collected by The Hong Kong University of Science and Technology (Guangzhou). The PhysDrive, comprises about 24 hours(1500K frames) of recordings from RGB camera, NIR camera, and an mmWave radar of 48 subjects. The dataset was designed to ontactless in-vehicle physiological sensing with dedicated consideration on various modality settings and driving factors, including Three vehicle types, four illumination conditions, three road conditions, and driver motions. PhysDrive is comprehensive with six synchronized ground truths (ECG, BVP, Respiration, HR, RR, and SpO2) and can be used in conjunction with the rPPG-toolbox.
Verbatim fromGitHub· Description from the repository README; licence: MIT
Key numbers
How it works
No model in the hub reports results on this dataset yet.
- Reference labels
- hr, resp rate, spo2
Models trained or evaluated on it
FAQ
How many subjects are in PhysDrive?
PhysDrive has data from 48 subjects, as listed in its research record.
What signals does PhysDrive include?
PhysDrive includes video, nir, mmwave, ECG, PPG and respiration data.
How do I get PhysDrive, and what is its licence?
It is available from https://github.com/WJULYW/PhysDrive-Dataset. Licence: Academic-only; commercial use banned; code MIT; raw data under DSA.
Cite
No citation metadata yet.
Cite the ChatPPG record
Cite the dated snapshot 2026-10 for a fixed, reproducible version. Please also cite the original publications listed above.
APA
ChatPPG. (2026). PhysDrive: dataset record (Version 2026-10) [Data set]. ChatPPG Research Hub. https://chatppg.com/research/datasets/physdrive
BibTeX
@misc{chatppg_dataset_physdrive,
author = {{ChatPPG}},
title = {{PhysDrive: dataset record}},
year = {2026},
month = oct,
version = {2026-10},
howpublished = {\url{https://chatppg.com/research/datasets/physdrive}},
note = {ChatPPG Research Hub, Snapshot 2026-10, https://chatppg.com/research/snapshots/2026-10}
}CITATION.cff
cff-version: 1.2.0 message: "If you use this resource, please cite it as below." type: dataset title: "PhysDrive: dataset record" authors: - name: "ChatPPG" url: "https://chatppg.com/research/datasets/physdrive" repository-artifact: "https://chatppg.com/research/snapshots/2026-10" version: "2026-10" date-released: "2026-10-24"
History
Record updated
Papers
- MS-rPPG: Multi-spectral State Space Model for Remote Photoplethysmography in Driver Monitoring SystemsShared topic: in vehicleICRA 2026 · 2026
DPIKAN: Decomposed Physics-Informed KANs for Physiological Signal Representation LearningShared topic: mmwaveICLR 2027 submission · 2026- PPG-Distill: Efficient Photoplethysmography Signals Analysis via Foundation Model DistillationShared topic: neurips 2025NeurIPS 2025 Workshop on Learning from Tim · 2025
