EEG-JEPA：用于脑电基础模型的结构化隐变量预测
EEG-JEPA: Structured Latent Prediction for EEG Foundation Models
浏览论文内容
中文总结 AI 辅助
本研究提出EEG-JEPA结构化隐变量预测框架，通过优化目标设计提升EEG基础模型性能，在EEG-FM-Bench等基准上实现了14任务、9任务的平衡准确率提升，为EEG基础建模提供了新方案。
中文摘要 AI 辅助
脑电（EEG）基础模型旨在从大规模未标注记录中学习可复用表征。常见的预训练策略是掩码波形重构，但直接对含噪EEG施加监督可能会促使模型恢复可预测的背景活动、采集效应及伪影，而非跨任务迁移的神经结构。这引发了一个核心问题：EEG基础模型应预测什么以学习可迁移的表征？我们提出EEG-JEPA，一种用于EEG基础建模的结构化隐变量预测框架。与重构掩码电压样本不同，掩码上下文编码器和预测器推断由指数移动平均目标编码器生成的上下文隐状态，该目标编码器观测完整输入。EEG-JEPA沿三个互补维度组织目标设计：目标内容指定预测的表征类型，目标支持通过神经拓扑感知多尺度电极-时间掩码（N-MET）在结构化电极-时间区域上指定预测位置，目标深度指定施加监督的编码器层。这些设计共同将EEG预训练从恢复缺失测量转向从结构化电极-时间上下文推断隐状态。我们通过受控客观比较、冻结多任务迁移和全微调评估EEG-JEPA：在相同主干、预训练语料库和训练时长下，EEG-JEPA相比CBraMod式掩码波形重构，将14任务冻结宏平均平衡准确率从40.49%提升至50.42%；多源续练进一步将该结果提升至52.94%，为EEG-FM-Bench上评估的EEG基础模型中的最高平均值；在协议匹配的全微调下，EEG-JEPA还将9任务平均平衡准确率从68.98%提升至70.65%。
英文摘要
Electroencephalography (EEG) foundation models aim to learn reusable representations from large-scale unlabeled recordings. A common pretraining strategy is masked waveform reconstruction, but applying supervision directly to noisy EEG may encourage models to recover predictable background activity, acquisition effects, and artifacts rather than neural structure that transfers across tasks. This raises a central question: what should an EEG foundation model predict to learn transferable representations? We introduce EEG-JEPA a structured latent-prediction framework for EEG foundation modeling. Rather than reconstructing masked voltage samples, a masked context encoder and predictor infer contextual latent states produced by an exponential-moving-average target encoder that observes the complete input. EEG-JEPA organizes target design along three complementary dimensions: target content specifies what representation is predicted, target support specifies where prediction occurs over structured electrode--time regions through Neurotopology-Aware Multi-scale Electrode-Temporal Masking (N-MET), and target depth specifies at which encoder layers supervision is applied. Together, these designs shift EEG pretraining from recovering missing measurements to inferring latent states from structured electrode--time context. We evaluate EEG-JEPA through controlled objective comparisons, frozen multitask transfer, and full fine-tuning. Under the same backbone, pretraining corpus, and training duration, EEG-JEPA improves the 14-task frozen macro balanced accuracy from 40.49% to 50.42% over CBraMod-style masked waveform reconstruction. Multi-source continuation further raises this result to 52.94%, the highest average among the EEG foundation models evaluated on EEG-FM-Bench. Under protocol-matched full fine-tuning, EEG-JEPA also improves the nine-task average balanced accuracy from 68.98% to 70.65%.
发表机构
- Tsinghua Laboratory of Brain and Intelligence, Tsinghua University（清华大学脑与智能实验室）
- School of Basic Medical Sciences, Tsinghua Medicine, Tsinghua University（清华大学清华医学基础医学院）
- School of Biomedical Engineering, Tsinghua Medicine, Tsinghua University（清华大学清华医学生物医学工程学院）
- Academy for Advanced Interdisciplinary Studies, Peking University（北京大学前沿交叉学科研究院）
- Department of Computer Science and Technology, Tsinghua University（清华大学计算机科学与技术系）
机构由 AI 辅助整理，请以论文原文为准。
