Leveraging Shared Prototypes for a Multimodal Pulse Motion Foundation Model
This addresses the challenge of modeling interconnected physiological processes for healthcare and medical research, though it appears incremental as it builds on existing self-supervised learning methods.
The paper tackled the problem of fragmented and non-generalizable embeddings in multimodal time-series data, particularly for biosignals like ECG and PPG, by proposing ProtoMM, a self-supervised learning framework that uses shared prototypes to anchor modalities in a common space, achieving state-of-the-art performance with improved interpretability.
Modeling multi-modal time-series data is critical for capturing system-level dynamics, particularly in biosignals where modalities such as ECG, PPG, EDA, and accelerometry provide complementary perspectives on interconnected physiological processes. While recent self-supervised learning (SSL) advances have improved unimodal representation learning, existing multi-modal approaches often rely on CLIP-style contrastive objectives that overfit to easily aligned features and misclassify valid cross-modal relationships as negatives, resulting in fragmented and non-generalizable embeddings. To overcome these limitations, we propose ProtoMM, a novel SSL framework that introduces a shared prototype dictionary to anchor heterogeneous modalities in a common embedding space. By clustering representations around shared prototypes rather than explicit negative sampling, our method captures complementary information across modalities and provides a coherent "common language" for physiological signals. In this work, we focus on developing a Pulse Motion foundation model with ProtoMM and demonstrate that our approach outperforms contrastive-only and prior multimodal SSL methods, achieving state-of-the-art performance while offering improved interpretability of learned features.