Separate audio into stems using various models
Extraction & Reconstruction for Efficient Speech Separation