arXiv cs.LG + cs.RO + cs.HC· Hanseul Choi, Jinyeong Park, Seongwon Jin, Sungho Park, Jibum Kim·· 16 小时前AI 评分58
CAMEL-CLIP：面向通道异构 EEG 的多模态文本对齐基础模型
CAMEL-CLIP: Channel-aware Multimodal Electroencephalography-text Alignment for Generalizable Brain Foundation Models
AI 导读
研究团队提出 CAMEL-CLIP，一种对通道组成和顺序变化更稳健的 EEG-文本对比学习基础模型。模型由基于通道解剖属性的位置编码、动态通道投影和通道级与样本级双层对比学习三部分组成。在线性探测下，该模型在多个数据集上取得所评估基线中最高的平均平衡准确率，并在不同通道配置间迁移时无需重新训练分类器。
正文 · 原文
Abstract:Electroencephalography (EEG) foundation models have shown promise for learning generalizable representations, yet they remain sensitive to channel heterogeneity, such as changes in channel composition or ordering. We propose channel-aware multimodal EEG-text alignment contrastive language image pretraining (CAMEL-CLIP), a contrastive EEG-text multimodal foundation model designed to be robust to heterogeneous channel configurations and widely applicable to diverse downstream tasks. CAMEL-CLIP combines three components: (1) channel attribute-based positional encoding, which identifies channels using their anatomical attributes and reference type; (2) dynamic channel projection, which preserves channel-specific information across varying numbers of EEG channels and combines textual reports with channel information to support channel-wise EEG-text matching; and (3) dual-level contrastive learning, which aligns EEG signals with text at both the individual-channel and whole-sample levels. Experimental results show that CAMEL-CLIP achieves the highest mean balanced accuracy among the evaluated baselines across multiple datasets under linear probing. CAMEL-CLIP also shows high classification performance when transferred between different channel configurations without classifier retraining.
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2603.13272 [cs.LG] |
| (or arXiv:2603.13272v2 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2603.13272 arXiv-issued DOI via DataCite |
Submission history
From: Jibum Kim [view email]
[v1]
Fri, 27 Feb 2026 05:05:08 UTC (3,096 KB)
[v2]
Sun, 25 Oct 2026 23:51:55 UTC (11,891 KB)
来源：arXiv cs.LG + cs.RO + cs.HC · arxiv.org