SPERA：球面先验 EEG 基础模型，几何与频率感知潜空间预测
原标题：SPERA: Spherical Prior EEG Foundation Model with Geometry- and Frequency-Aware Latent Prediction
AI 摘要
研究团队提出 SPERA，一种基于球面先验的 EEG 基础模型，采用联合嵌入预测架构（JEPA）在潜空间进行预测，并引入 Legendre 多项式空间先验编码头皮电极几何结构。模型在来自 106 个数据集、29,048 名受试者、约 8 万小时 EEG 数据上预训练，在临床、认知和 BCI 等九项下游任务上取得最高平均平衡准确率，并展现出较强的参数效率和跨记录条件鲁棒性。
正文节选
SPERA: Spherical Prior EEG Foundation Model with Geometry- and Frequency-Aware Latent Prediction Abstract Electroencephalography (EEG) provides a non-invasive measure of ongoing neural activity, but building general-purpose EEG models remains challenging due to the heterogeneity of subjects, devices, and electrode montages. Existing EEG foundation models predominantly rely on reconstruction-based objectives defined on the observed signal, which contains both neural and non-neural components. We