Enregistré dans:
Détails bibliographiques
Auteurs principaux: Liu, Dingkun, Chen, Yuheng, Chen, Zhu, Cui, Zhenyao, Wen, Yaozhi, An, Jiayu, Luo, Jingwei, Wu, Dongrui
Format: Preprint
Publié: 2026
Sujets:
Accès en ligne:https://arxiv.org/abs/2601.17883
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866908815282143232
author Liu, Dingkun
Chen, Yuheng
Chen, Zhu
Cui, Zhenyao
Wen, Yaozhi
An, Jiayu
Luo, Jingwei
Wu, Dongrui
author_facet Liu, Dingkun
Chen, Yuheng
Chen, Zhu
Cui, Zhenyao
Wen, Yaozhi
An, Jiayu
Luo, Jingwei
Wu, Dongrui
contents Electroencephalography (EEG) foundation models have recently emerged as a promising paradigm for brain-computer interfaces (BCIs), aiming to learn transferable neural representations from large-scale heterogeneous recordings. Despite rapid progresses, there lacks fair and comprehensive comparisons of existing EEG foundation models, due to inconsistent pre-training objectives, preprocessing choices, and downstream evaluation protocols. This paper fills this gap. We first review 50 representative models and organize their design choices into a unified taxonomic framework including data standardization, model architectures, and self-supervised pre-training strategies. We then evaluate 12 open-source foundation models and competitive specialist baselines across 13 EEG datasets spanning nine BCI paradigms. Emphasizing real-world deployments, we consider both cross-subject generalization under a leave-one-subject-out protocol and rapid calibration under a within-subject few-shot setting. We further compare full-parameter fine-tuning with linear probing to assess the transferability of pre-trained representations, and examine the relationship between model scale and downstream performance. Our results indicate that: 1) linear probing is frequently insufficient; 2) specialist models trained from scratch remain competitive across many tasks; and, 3) larger foundation models do not necessarily yield better generalization performance under current data regimes and training practices.
format Preprint
id arxiv_https___arxiv_org_abs_2601_17883
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle EEG Foundation Models: Progresses, Benchmarking, and Open Problems
Liu, Dingkun
Chen, Yuheng
Chen, Zhu
Cui, Zhenyao
Wen, Yaozhi
An, Jiayu
Luo, Jingwei
Wu, Dongrui
Machine Learning
Computer Vision and Pattern Recognition
Electroencephalography (EEG) foundation models have recently emerged as a promising paradigm for brain-computer interfaces (BCIs), aiming to learn transferable neural representations from large-scale heterogeneous recordings. Despite rapid progresses, there lacks fair and comprehensive comparisons of existing EEG foundation models, due to inconsistent pre-training objectives, preprocessing choices, and downstream evaluation protocols. This paper fills this gap. We first review 50 representative models and organize their design choices into a unified taxonomic framework including data standardization, model architectures, and self-supervised pre-training strategies. We then evaluate 12 open-source foundation models and competitive specialist baselines across 13 EEG datasets spanning nine BCI paradigms. Emphasizing real-world deployments, we consider both cross-subject generalization under a leave-one-subject-out protocol and rapid calibration under a within-subject few-shot setting. We further compare full-parameter fine-tuning with linear probing to assess the transferability of pre-trained representations, and examine the relationship between model scale and downstream performance. Our results indicate that: 1) linear probing is frequently insufficient; 2) specialist models trained from scratch remain competitive across many tasks; and, 3) larger foundation models do not necessarily yield better generalization performance under current data regimes and training practices.
title EEG Foundation Models: Progresses, Benchmarking, and Open Problems
topic Machine Learning
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2601.17883