SPOTR: Spatio-temporal Pooling One-Token Reconstruction for Universal Physiological Signal Self-supervised Learning
For researchers and practitioners in physiological signal analysis, SPOTR provides a universal SSL method that achieves strong performance across multiple modalities with high efficiency, addressing limitations of existing methods in handling heterogeneous datasets and computational costs.
SPOTR introduces a compress-reconstruct pretraining framework with a single-token global bottleneck for physiological signals, outperforming the strongest baseline under linear probing by improving average AUC by 18.49%, 21.71%, 17.86%, and 4.64% across EEG, iEEG, ECG, and PPG datasets, while reducing latency by ~78% and peak GPU memory by ~52%.
Physiological signals such as EEG, ECG, and PPG are widely used in clinical monitoring. Recent self-supervised learning (SSL) methods offer an attractive way to leverage unlabeled recordings, yet they still fall short in practice. In particular, current SSL methods struggle across heterogeneous datasets, often distorting clinically meaningful structures or learning shortcuts from temporal and cross-channel redundancy. Consequently, existing SSL methods often deliver limited performance under linear probing, a lightweight adaptation setting that better matches real-world medical scenarios. Moreover, most Transformer-based SSL models encode a flattened spatiotemporal token sequence, incurring high computation and memory cost, and are typically developed within a single modality. To address these limitations, we present SPOTR (Spatio-temporal Pooling One-Token Reconstruction), a compress-reconstruct pretraining framework that introduces a single-token global bottleneck for physiological signals. SPOTR compresses each waveform into a single-token representation and reconstructs the signal conditioned only on this representation. Meanwhile, SPOTR introduces an efficient spatio-temporal compaction module to reduce computation and memory cost. Pretrained on 20 datasets spanning EEG, iEEG, ECG, and PPG, SPOTR consistently outperforms the strongest baseline under linear probing, improving average AUC by 18.49%, 21.71%, 17.86%, and 4.64%, respectively. Compared with a representative general-purpose time-series foundation model, SPOTR achieves around 78% lower latency and 52% lower peak GPU memory on average. The code can be found at https://github.com/5GYYYYY/SPOTR.