STAMP: Spatial-Temporal Adapter with Multi-Head Pooling
This work addresses the problem of efficiently modeling EEG data for clinical classification tasks, offering a flexible and parameter-efficient solution, though it is incremental as it builds on existing foundation models.
The paper tackled the lack of comparative analysis between EEG-specific foundation models and general time series foundation models on EEG tasks by introducing STAMP, a lightweight adapter that leverages general TSFM embeddings to achieve performance comparable to state-of-the-art EEGFMs on 8 benchmark datasets.
Time series foundation models (TSFMs) pretrained on data from multiple domains have shown strong performance on diverse modeling tasks. Various efforts have been made to develop foundation models specific to electroencephalography (EEG) data, which records brain electrical activity as time series. However, no comparative analysis of EEG-specific foundation models (EEGFMs) versus general TSFMs has been performed on EEG-specific tasks. We introduce a novel Spatial-Temporal Adapter with Multi-Head Pooling (STAMP), which leverages univariate embeddings produced by a general TSFM, implicitly models spatial-temporal characteristics of EEG data, and achieves performance comparable to state-of-the-art EEGFMs. A comprehensive analysis is performed on 8 benchmark datasets of clinical tasks using EEG for classification, along with ablation studies. Our proposed adapter is lightweight in trainable parameters and flexible in the inputs it can accommodate, supporting easy modeling of EEG data using TSFMs.