Voice Separation with an Unknown Number of Multiple Speakers
This addresses the challenge of voice separation in audio processing for applications like speech recognition, with incremental improvements in handling unknown speaker counts.
The paper tackles the problem of separating mixed audio with multiple simultaneous speakers, achieving state-of-the-art performance that significantly outperforms existing methods, especially for more than two speakers.
We present a new method for separating a mixed audio sequence, in which multiple voices speak simultaneously. The new method employs gated neural networks that are trained to separate the voices at multiple processing steps, while maintaining the speaker in each output channel fixed. A different model is trained for every number of possible speakers, and the model with the largest number of speakers is employed to select the actual number of speakers in a given sample. Our method greatly outperforms the current state of the art, which, as we show, is not competitive for more than two speakers.