Fetching the paper…

VisualVoice: Audio-Visual Speech Separation with Cross-Modal Consistency · Around