Reproduce “Complementarity-Driven Dual-Teacher Hierarchical Alignment for Zero-Shot EEG-to-Image Retrieval” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Lukun Wu, Jie Li, Ziqi Ren, Ruiyang Xia, Yan Gao, Xinbo Gao
DTHA improves EEG-to-image retrieval by aligning EEG signals separately with multi-blur CLIP and shallow DINO visual teachers, then fusing their similarity scores. On the 200-way THINGS-EEG intra-subject test, it reaches 92.4% Top-1, averaged across 10 participants and three seeds, 12.4 percentage points above LVBP. However, leave-one-subject-out Top-1 is 25.6%, showing cross-subject retrieval remains much lower.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.