Reproduce “Spectral Considerations on Unsupervised Time-domain EEG Representation Learning with Deep Networks” 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
David Aquilué-Llorens, Abel Puigseslloses, Angel Blanco, Claire Braboszcz, Aureli Soria-Frisch
The authors examine how EEG autoencoders learn to reconstruct frequency components and whether compressed embeddings retain synthetic cognitive-group differences. In controlled signals, declining amplitudes across frequency led LSTMVAE, VAEEG, and Transformer models to recover higher frequencies later. In a synthetic β/γ task, embeddings of at least 100 dimensions supported group classification for LSTMVAE and Transformer, not VAEEG; this is not clinical diagnosis.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.