Reproduce “BrainBind: Bridging Synchronized EEG and fMRI through Long-Form Naturalistic Video Representations” 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

Shixiao Wang, Rining Wu, Jian K. Liu

BrainBind aligns synchronized EEG and fMRI from video viewing with frozen video embeddings for bidirectional retrieval. Neural encoders map each signal into this shared reference space, and Joint-Dual averages modality embeddings. On CineBrain’s random split, Joint-Dual leads fMRI in label-aware R@1, while fMRI leads in MRR and nDCG@10. This system-level contrast does not isolate fusion; retrieval drops on episode-disjoint evaluation.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.