· – High frequency magnitude spectrogram reconstruction for music mixtures using convolutional autoencoders

Music Bandwidth Expansion

autoencoders in frequency domain yet again ... Read More ›

· – Review of interpretability in machine learning


Inside the black box ... Read More ›

· – Telefonica internship

Catastrophic forgetting

Hard attention on the task ... Read More ›


Deep learning source separation for hip hop and classical music

Source separation papers at ISMIR conference and MML workshop... Read More ›


Sound and Music Computing conference

Paper on generating training data for deep learning source separation methods... Read More ›