ERP-XTTN: Interpretable prototype cross-attention for ERP classification

Arxiv CS.LG (machine learning)

Proposes ERP-XTTN, a prototype-guided cross-attention mechanism for cross-subject ERP classification without calibration. Aims to provide competitive performance with interpretable attention maps highlighting relevant EEG features.

interpretability

Arxiv CS.LG (machine learning)·Charlotte Genevier WymanLeanne Hirshfield

8 days ago