Audio Source Separation using Low Latency Neural Network

This reposiory contains the code for our course project for Machine Learning (CS419) at IIT Bombay. We have used the PyTorch library to construct a neural network to separate instruments from a music file. We have implemented the paper "Monoaural Audio Source Separation Using Deep Convolutional Neural Networks", along with a few modifications and experiments inspired by other papers.

Team Members

• [16D100012] Sarthak Consul (@SConsul)
• [160110085] Archiki Prasad (@archiki)
• [16D070001] Parthasarathi Khirwadkar (@kparth98)
• [16D100001] Deepak Gopalan (@DeepakGopalan)

Bibliography

• [1] Pritish Chandna, M. Miron, Jordi Janer, and Emilia G´omez. Monoaural audio source separation using deep convolutional neural networks. In 13th International Conference on Latent Variable Analysis and Signal Separation (LVAICA2017), 02/2017 2017
• [2] E. Vincent, R. Gribonval, and C. Fevotte. Performance measurement in blind audio source separation. IEEE Transactions on Audio, Speech, and Language Processing, 14(4):1462–1469, July 2006
• [3] Chao Peng, Xiangyu Zhang, Gang Yu, Guiming Luo, and Jian Sun. Large kernel matters - improve semantic segmentation by global convolutional network. CoRR, abs/1703.02719, 2017