1 paper
Héctor Martel, Julius Richter, Kai Li +2
We propose Audio-Visual Lightweight ITerative model (AVLIT), an effective and lightweight neural network that uses Progressive Learning (PL) to perform audio-visual speech separati…