Temporal Feature Extractors in EEG Foundation Models: A Controlled Comparison Including a Pretrained Time-Series Model
2026-06-29 • Artificial Intelligence
Artificial Intelligence
AI summaryⓘ
The authors studied how different methods to extract time-based features from brain wave recordings (EEG) affect the quality of learned representations. They compared a simple linear method, a convolutional encoder, and a pretrained general time-series model called MOMENT. Their tests on two tasks—recognizing imagined movements and emotions—showed that simple methods work well for movement-related data, but more complex models help with emotion data. They also found that using a pretrained model not specifically made for EEG can still be useful for extracting time features in EEG data.
Electroencephalography (EEG)Temporal feature extractionTime-series foundation models (TSFMs)Convolutional encoderMotor imageryEmotion recognitionRepresentation learningPretrained modelsFrozen feature extractor
Authors
Ayşe Betül Yüce, Chris Joey Leffler, Sarun Varghese, Myra Spiliopoulou, Sebastian Stober
Abstract
Electroencephalography (EEG) foundation models aim to learn generalizable representations from large-scale brain recordings. However, the role of temporal feature extractors and whether pretrained time-series foundation models (TSFMs) can be effectively transferred to this setting remains underexplored. We conduct a controlled comparison of three temporal feature extraction strategies, including a linear baseline, a convolutional encoder, and a frozen pretrained TSFM (MOMENT), within a unified EEG foundation model. We evaluate their impact on representation quality using two downstream tasks: motor imagery and emotion recognition. Results reveal different trends across the evaluated benchmarks. On the motor imagery dataset, simple temporal representations perform competitively, whereas the emotion dataset benefits from richer temporal modeling. Although not specifically adapted to EEG, the pretrained TSFM serves as an effective temporal feature extractor, suggesting that general-purpose time-series representations can be transferred as frozen temporal feature extractors within EEG foundation models.