Taketo Akama

Unpaired data can unlock superior dance-music alignment, achieving better results than traditional paired data methods.

Separating acoustic and expectation-related representations in neural networks unlocks better EEG-based music identification, outperforming even strong ensembles.