Models, datasets
& papers.

compute-optimal physiological waveform FM family (20M–2.1B)

We investigate the scaling laws and compute-optimal training of physiological waveform foundation models (FMs). We train Aether, a family of over one hundred FMs ranging from 20M…

Why this family: The catalog reports learning across multiple signals or derived sensor channels. Training modalities may differ from inference inputs.

Why this family: No reviewed primary-role assignment. See the source record for the reported method.

AnyMo models human motion from IMUs placed anywhere on the body. It simulates IMU signals over dense body-surface positions, pretrains a graph encoder on them, turns multi-position IMU data into full-body motion tokens and aligns those tokens with an LLM. Across 14 unseen datasets it improved zero-shot activity recognition accuracy by 11.7%, and it also improved IMU-text retrieval and motion captioning.

As wearable and mobile devices become increasingly embedded in daily life, they offer a practical way to continuously sense human motion in the wild. But inertial signals are…

Why this family: The catalog reports language alignment, signal-to-text interpretation or a sensor-grounded language system.

ECG-guided CLIP-style PPG FM (PPG encoder aligned to frozen ECGFounder) over ~110k hours paired clinical PPG+ECG; inference is PPG-only.

Photoplethysmography (PPG) is widely used as a non-invasive and accessible modality for continuous health monitoring. However, despite being a peripheral hemodynamic signal…

Why this family: PPG encoder aligned to an ECG teacher during training; the record explicitly reports PPG-only inference.

This Apple model is an accelerometer foundation model trained by distilling knowledge from a PPG encoder. It used 20 million minutes of unlabelled data from about 172K participants in the Apple Heart and Movement Study. Accelerometer embeddings could retrieve the matching PPG embedding with 99.2% top-1 accuracy, and the distilled encoders beat accelerometer-only encoders by 23% to 49% on health targets.

Modern wearable devices can conveniently record various biosignals in the many different environments of daily living, enabling a rich view of individual health. However, not all…

Why this family: An accelerometer student encoder distilled from a PPG teacher; this does not require both signals at inference.

Wearable devices record physiological and behavioral signals that can improve health predictions. While foundation models are increasingly used for such predictions, they have…

Why this family: Grouped by its reported signal-representation or task-encoding role. See the record for the training setup and supported inputs.

Reported architecture: State space

Apple ICLR 2024 large-scale PPG/ECG FMs with participant-level contrastive pairs on ~141k AHMS participants — no downloadable checkpoint on official page.

Tracking biosignals is crucial for monitoring wellness and preempting the development of severe medical conditions. Today, wearable devices can conveniently record various…

Why this family: A study of PPG and ECG encoders; listing both modalities does not establish joint-input inference.

delay-aware beat-resolved ECG–PPG cardiovascular FM

Electrocardiography and photoplethysmography record the same heartbeat at two ends of the cardiovascular system, separated by a beat-to-beat delay that reflects ejection dynamics…

Why this family: The catalog reports learning across multiple signals or derived sensor channels. Training modalities may differ from inference inputs.

Bio-PM is a self-supervised model for wrist accelerometer data that splits motion into movement segments based on the submovement theory of motor control. A Transformer encoder is pretrained to reconstruct masked segments, which focuses it on how movements are organized rather than on raw waveform shape. Pretrained on about 28,000 hours from 11,000 NHANES participants, it beat strong wearable SSL baselines on six subject-disjoint activity recognition benchmarks.

Wearable accelerometers enable large-scale health monitoring, yet learning robust human-activity representations has been constrained by scarce labeled data. While self-supervised…

Why this family: Grouped by its reported signal-representation or task-encoding role. See the record for the training setup and supported inputs.

Reported architecture: Transformer

BLPM (Brain Latent Predictive Model) is an EEG-language foundation model that treats EEG decoding tasks as predicting continuous semantic embeddings. Its CELP encoder learns by predicting latent targets, and a multi-query module aligns EEG representations with text in a shared space. It generalized consistently across several benchmarks.

Recent advances in EEG foundation models have demonstrated the potential of large-scale pretraining to enable generalizable neural decoding across subjects, recording…

Why this family: The catalog reports language alignment, signal-to-text interpretation or a sensor-grounded language system.

Patient-level PPG↔EHR contrastive foundation-style encoder (KDD 2026). Anchors long PPG windows to LLM-distilled clinical semantics from MC-MED ED data; reports strong AF/HR/RR/BP transfer vs SiamQuality / GPT-PPG / PaPaGei.

Photoplethysmography (PPG) plays a central role in wearable health monitoring and clinical decision support. Yet existing approaches to universal PPG representation learning…

Why this family: The catalog reports language alignment, signal-to-text interpretation or a sensor-grounded language system.

Foundation Model for Capacitive Touch Sensing

Why this family: Announced capacitive-touch model; insufficient detail to assign a primary role.

CardioFM is a self-supervised foundation model that encodes ECG lead II and PPG together, using cross-modal attention and residual vector quantization. It reached an F1 of 0.86 for cardiovascular disease classification on PTB-XL and estimated QT interval with a mean error of 20.2 ms and pulse arrival time with 22.7 ms. Its embeddings also predict age, sex and BMI without diagnostic labels.

Electrocardiography (ECG) and photoplethysmography (PPG) arise from the same heartbeat and are routinely co-acquired at every monitored bedside, yet no foundation model jointly…

Why this family: The catalog reports learning across multiple signals or derived sensor channels. Training modalities may differ from inference inputs.

CardioState-JEPA is a cardiac foundation model that learns one shared representation across ECG, PPG and heart sounds (PCG). Each signal is mapped into a common token space and fed to a single Transformer that learns by predicting masked latent cardiac states. As a frozen encoder across 25 tasks, it improved PPG classification by 8.2 AUROC points, murmur detection by 18.8 and ECG classification by 15.5 over the best baselines.

Electrocardiography (ECG), photoplethysmography (PPG), and phonocardiography (PCG) provide complementary views of the same cardiac cycle, yet existing cardiac foundation models…

Why this family: The catalog reports learning across multiple signals or derived sensor channels. Training modalities may differ from inference inputs.

Reported architecture: Transformer

CLEF is a foundation model for single-lead ECG trained with contrastive learning guided by patient risk scores derived from clinical metadata. It was pretrained on 12-lead ECGs from 161K patients in MIMIC-IV-ECG and tested on 18 classification and regression tasks across 7 held-out datasets. It improved average AUROC by at least 2.6% and cut regression error by at least 3.2% compared with self-supervised baselines, and weights are released in small, base and large sizes.

The electrocardiogram (ECG) is a key diagnostic tool in cardiovascular health. Single-lead ECG recording is integrated into both clinical-grade and consumer wearables. While…

Why this family: Grouped by its reported signal-representation or task-encoding role. See the record for the training setup and supported inputs.