基于原型的PPG与加速度计信号自监督多模态学习
Prototype-based Self-Supervised Multimodal Learning for PPG and Accelerometry Signals
AI总结：
针对PPG与加速度计等生物信号的多模态学习局限，提出基于原型的SSL框架ProtoMM，开发Pulse-Motion基础模型，性能优于现有对比式及多模态SSL方法。
摘要：
对多模态时间序列数据进行建模对于捕捉系统级动态至关重要，尤其是在生物信号领域，ECG、PPG、EDA和加速度计等模态为相互关联的生理过程提供了互补视角。尽管近期自监督学习（SSL）的进展提升了单模态表示学习能力，但现有多模态方法常依赖CLIP式对比目标，该目标易过拟合于易对齐特征，并将有效的跨模态关系误判为负样本，导致嵌入表示碎片化且泛化性差。为克服这些局限，我们提出ProtoMM——一种新型SSL框架，引入共享原型字典将异构模态锚定在公共嵌入空间中。我们的方法通过围绕共享原型对表示进行聚类，而非采用显式负采样，可捕捉跨模态互补信息，并为生理信号提供连贯的“通用语言”。本研究聚焦于利用ProtoMM开发Pulse-Motion基础模型，实验表明，我们的方法仅在对比学习及现有多模态SSL方法中表现更优，在强现有基准中取得最佳性能，同时为分析和解释学习到的特征提供额外实用性。
英文摘要：
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 the best performance among strong prior baselines while offering additional utility for analyzing and interpreting learned features.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
