mask separation approach

This article introduces a new method for audio-visual speech separation, called RTFS-Net. This method operates in the time-frequency domain and uses a multi-layered RNN to independently model and capture the…