ddcas / VocalSeparation

Using Deep Learning for singing voice separation - Project for the course DT2119 Speech and Speaker Recognition offered by KTH in 2018

MusicToLyrics

The project in inspired by [1].

Execution instructions:

1. Download MIR-1K dataset from https://sites.google.com/site/unvoicedsoundseparation/mir-1k
2. Execute create_dataset.py to split and preprocess data
3. Execute train.py to train the network
4. Execute eval.py to evaluate the network

[1] P.-S. Huang, M. Kim, M. Hasegawa-Johnson, and P. Smaragdis, “Singing-voice separation from monaural recordings using deep recurrent neural networks.,” in ISMIR, 2014, pp. 477–482.