Train on noisy audio

[soroushhashemifar]

I wanted to know what happens if I train the unsupervised model on unlabelled audios with music and background noise. Will that still work?