REVIEW 18 cited by
Scaling Wearable Foundation Models
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
read the original abstract
Wearable sensors have become ubiquitous thanks to a variety of health tracking features. The resulting continuous and longitudinal measurements from everyday life generate large volumes of data; however, making sense of these observations for scientific and actionable insights is non-trivial. Inspired by the empirical success of generative modeling, where large neural networks learn powerful representations from vast amounts of text, image, video, or audio data, we investigate the scaling properties of sensor foundation models across compute, data, and model size. Using a dataset of up to 40 million hours of in-situ heart rate, heart rate variability, electrodermal activity, accelerometer, skin temperature, and altimeter per-minute data from over 165,000 people, we create LSM, a multimodal foundation model built on the largest wearable-signals dataset with the most extensive range of sensor modalities to date. Our results establish the scaling laws of LSM for tasks such as imputation, interpolation and extrapolation, both across time and sensor modalities. Moreover, we highlight how LSM enables sample-efficient downstream learning for tasks like exercise and activity recognition.
Forward citations
Cited by 18 Pith papers
-
Detecting Drunk Driving Using Off-the-Shelf Smartwatches
Machine learning models using smartwatch data from a 54-participant test-track study detect alcohol-impaired driving with participant-averaged AUROC of 0.88 for any intoxication and 0.86 above 0.05 g/dL.
-
Model Merging as Probabilistic Inference in Fine-Tuning Parameter Space
Model merging is cast as PoE inference with EBM experts, revealing Gaussian assumptions in prior work and proposing convergent Cauchy experts that improve empirical performance.
-
BCG-FM: A Foundation Model for Ambient Cardiac Health Sensing
BCG-FM, the first foundation model for ambient BCG, achieves 3.26-year MAE on biological age estimation and discriminates 15 health conditions using frozen embeddings from participant-level contrastive pretraining on ...
-
OpenWatch: A Multimodal Benchmark for Hand Gesture Recognition on Smartwatches
OpenWatch provides the first open multimodal smartwatch gesture dataset and benchmark, with MixToken and NormWear-Lora methods reaching 90% F1-score using 223k parameters versus 66% for 136M-parameter foundation models.
-
Physiology-Aware Masked Cross-Modal Reconstruction for Biosignal Representation Learning
xMAE pretrains biosignal representations via masked cross-modal reconstruction of temporally ordered signals like ECG and PPG, outperforming baselines on 15 of 19 downstream tasks including cardiovascular prediction a...
-
A Diffusion-Model Subpopulation Digital Twin for Mobile Health Deployment: A Case Study on the HeartSteps Intervention
A pre-train/fine-tune/calibrate diffusion-model pipeline produces subpopulation digital twins that out-reproduce simpler simulators on temporal and between-participant structure in a HeartSteps replay.
-
Signal or Noise? Understanding Generative Models for Real-World Sensor Time Series
Across 14 sensor generation settings, flow-matching models are the strongest overall baseline, while demographic conditioning, time-frequency modeling, and moderate synthetic augmentation improve hard regimes and down...
-
TimeSRL: Generalizable Time-Series Behavioral Modeling via Semantic RL-Tuned LLMs -- A Case Study in Mental Health
TimeSRL uses semantic abstractions from time-series data optimized via reinforcement learning to achieve better cross-dataset generalization than standard ML or LLM baselines in mental health prediction.
-
Emergent Symbolic Structure in Health Foundation Models: Extraction, Alignment, and Cross-Modal Transfer
Health foundation model embeddings contain an interpretable symbolic organization shared across modalities that supports cross-domain transfer without joint training.
-
OSF: On Pre-training and Scaling of Sleep Foundation Models
Channel-masked self-supervised pretraining on a 166,500-hour multi-source sleep corpus yields OSF, which generalizes better to missing channels and scales with data and model size.
-
Human Centric Embodied Intelligence for Soft Wearable Robotics
A narrative review introducing Human-Centric Embodied Intelligence (HCEI) and the PCAA framework as organizing principles for soft wearable robotics.
-
WEQA: Wearable hEalth Question Answering with Query-Adaptive Agentic Reasoning
WEQA proposes a query-adaptive agent framework combining LLMs with wearable data tools, achieving 24% higher accuracy than baselines on a benchmark from four open datasets, with gains in expert-rated usefulness.
-
Gravity-Aware Hierarchical Routing for Lightweight SensorLLM on Human Activity Recognition
Introduces a lightweight gravity-aware routing head that improves macro-F1 on static classes in compressed SensorLLM for human activity recognition on the MHealth dataset.
-
Emergent Symbolic Structure in Health Foundation Models: Extraction, Alignment, and Cross-Modal Transfer
Independently trained PPG and accelerometer health foundation models share a linearly alignable subspace in which health-condition classifiers transfer with >95% of in-domain AUC.
-
Physical Foundation Models: Fixed hardware implementations of large-scale neural networks
Physical Foundation Models are fixed physical hardware realizations of foundation-scale neural networks that compute via inherent material dynamics, potentially delivering orders-of-magnitude gains in energy efficienc...
-
Wearable AI in the Era of Large Sensor Models
Large Sensor Models trained on large-scale multimodal wearable data can provide a scalable, general framework for wearable AI by learning transferable representations across modalities and tasks.
-
Foundation Models Defining A New Era In Sensor-based Human Activity Recognition: A Survey And Outlook
The survey organizes foundation models for sensor-based HAR into a lifecycle taxonomy and identifies three trajectories: HAR-specific models from scratch, adaptation of general time-series models, and integration with...
-
Cross-device Zero-shot Label Transfer via Alignment of Time Series Foundation Model Embeddings
A framework using adversarial alignment of frozen time-series foundation model embeddings transfers labels to a simulated target domain, but the target is synthetic and no real consumer device data is tested.
Discussion (0). Sign in to comment.