STAMP: Spatial-Temporal Adapter with Multi-Head Pooling

Brad Shook, Abby Turner, Jieshi Chen, Michał Wiliński, Mononito Goswami, Jonathan Elmer, Artur Dubrawski

Keywords: time series foundation models (TSFM), electroencephalography (EEG) data, spatial-temporal adapter, EEG foundation models (EEGFM)

TL;DR: We introduce a novel Spatial-Temporal Adapter with Multi-Head Pooling (STAMP) which is used on top of general-purpose time series foundation models to model electroencephalography (EEG) data.

Track: Proceedings

Abstract: 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 **S**patial-**T**emporal **A**dapter with **M**ulti-Head **P**ooling (**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.

General Area: Models and Methods

Specific Subject Areas: Foundation Models, Time Series, Supervised Learning

Data And Code Availability: Yes

Ethics Board Approval: No

Entered Conflicts: I confirm the above

Anonymity: I confirm the above

Code URL: https://github.com/autonlab/STAMP

Submission Number: 177