Guardado en:
Detalles Bibliográficos
Autores principales: Cao, Peng, Yang, Zhijian, Liu, Tennison, Wang, Jonathan, Wu, Jiang, Proszewska, Magdalena, Pillai, Arvind, Gao, Mingwu, Farjadian, Amir, Cai, Lawrence, Blanchard, Emily, McDuff, Daniel, Rudrapatna, Pramod, Thompson, Matthew, Pathak, Anupam, Malhotra, Mark, Patel, Shwetak, Katabi, Dina, Di Achille, Paolo, Poh, Ming-Zher
Formato: Preprint
Publicado: 2026
Materias:
Acceso en línea:https://arxiv.org/abs/2605.09173
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866913107471761408
author Cao, Peng
Yang, Zhijian
Liu, Tennison
Wang, Jonathan
Wu, Jiang
Proszewska, Magdalena
Pillai, Arvind
Gao, Mingwu
Farjadian, Amir
Cai, Lawrence
Blanchard, Emily
McDuff, Daniel
Rudrapatna, Pramod
Thompson, Matthew
Pathak, Anupam
Malhotra, Mark
Patel, Shwetak
Katabi, Dina
Di Achille, Paolo
Poh, Ming-Zher
author_facet Cao, Peng
Yang, Zhijian
Liu, Tennison
Wang, Jonathan
Wu, Jiang
Proszewska, Magdalena
Pillai, Arvind
Gao, Mingwu
Farjadian, Amir
Cai, Lawrence
Blanchard, Emily
McDuff, Daniel
Rudrapatna, Pramod
Thompson, Matthew
Pathak, Anupam
Malhotra, Mark
Patel, Shwetak
Katabi, Dina
Di Achille, Paolo
Poh, Ming-Zher
contents Wearable sensors enable the continuous acquisition of high-resolution physiological waveforms, such as photoplethysmography and accelerometry, under free-living conditions. However, inferring health-related phenotypes from these signals presents significant challenges due to high sampling frequencies, multimodal dependencies, and extreme sequence lengths (e.g., weeks of recordings), compounded by a scarcity of ground-truth labels. To address these challenges, existing self-supervised learning (SSL) methodologies typically follow two paradigms: (1) learning rich morphological representations from short waveform segments while collapsing longitudinal dynamics through simple aggregation, or (2) modeling behavioral patterns from coarse, hand-crafted features (e.g. heart rate, step counts) spanning longer horizons but foregoing subtle, predictive signatures in raw waveforms. To bridge this gap, we propose WavesFM, a foundation model utilizing a two-stage SSL framework for longitudinal physiological data. Specifically, we decompose the learning problem into two stages: first, a segment-level encoder is pretrained to extract local embeddings from short waveforms; subsequently, a temporal encoder is trained to model the sequence of these embeddings across a multi-day horizon. This hierarchical approach overcomes the computational complexity of high-resolution, long-sequence data, allowing the overall model to capture both local signal semantics and the complex circadian and inter-day variations governing physiological dynamics. Pretrained on over 6.8M hours (N=324k individuals) of recordings for the first stage and 5.3M hours (N=10k) for the second stage, WavesFM demonstrates superior performance across 58 diverse tasks spanning demographics, lifestyle, health conditions, and medications.
format Preprint
id arxiv_https___arxiv_org_abs_2605_09173
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle WavesFM: Hierarchical Representation Learning for Longitudinal Wearable Sensor Waveforms
Cao, Peng
Yang, Zhijian
Liu, Tennison
Wang, Jonathan
Wu, Jiang
Proszewska, Magdalena
Pillai, Arvind
Gao, Mingwu
Farjadian, Amir
Cai, Lawrence
Blanchard, Emily
McDuff, Daniel
Rudrapatna, Pramod
Thompson, Matthew
Pathak, Anupam
Malhotra, Mark
Patel, Shwetak
Katabi, Dina
Di Achille, Paolo
Poh, Ming-Zher
Machine Learning
Artificial Intelligence
Wearable sensors enable the continuous acquisition of high-resolution physiological waveforms, such as photoplethysmography and accelerometry, under free-living conditions. However, inferring health-related phenotypes from these signals presents significant challenges due to high sampling frequencies, multimodal dependencies, and extreme sequence lengths (e.g., weeks of recordings), compounded by a scarcity of ground-truth labels. To address these challenges, existing self-supervised learning (SSL) methodologies typically follow two paradigms: (1) learning rich morphological representations from short waveform segments while collapsing longitudinal dynamics through simple aggregation, or (2) modeling behavioral patterns from coarse, hand-crafted features (e.g. heart rate, step counts) spanning longer horizons but foregoing subtle, predictive signatures in raw waveforms. To bridge this gap, we propose WavesFM, a foundation model utilizing a two-stage SSL framework for longitudinal physiological data. Specifically, we decompose the learning problem into two stages: first, a segment-level encoder is pretrained to extract local embeddings from short waveforms; subsequently, a temporal encoder is trained to model the sequence of these embeddings across a multi-day horizon. This hierarchical approach overcomes the computational complexity of high-resolution, long-sequence data, allowing the overall model to capture both local signal semantics and the complex circadian and inter-day variations governing physiological dynamics. Pretrained on over 6.8M hours (N=324k individuals) of recordings for the first stage and 5.3M hours (N=10k) for the second stage, WavesFM demonstrates superior performance across 58 diverse tasks spanning demographics, lifestyle, health conditions, and medications.
title WavesFM: Hierarchical Representation Learning for Longitudinal Wearable Sensor Waveforms
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2605.09173