Single Channel Audio Source Separation using Convolutional Denoising Autoencoders

Abstract

Deep learning techniques have been used recently to tackle the audio source separation problem. In this work, we propose to use deep fully convolutional denoising autoencoders (CDAEs) for monaural audio source separation. We use as many CDAEs as the number of sources to be separated from the mixed signal. Each CDAE is trained to separate one source and treats the other sources as background noise. The main idea is to allow each CDAE to learn suitable spectral-temporal filters and features to its corresponding source. Our experimental results show that CDAEs perform source separation slightly better than the deep feedforward neural networks (FNNs) even with fewer parameters than FNNs.

Accepted at GlobalSIP 2017 and the final version is available at http://epubs.surrey.ac.uk/841860/

References in corpus (4)

• Theano: new features and speed improvements
• A Fully Convolutional Deep Auditory Model for Musical Chord Recognition
• Convolutional Recurrent Neural Networks for Music Classification
• Convolutional Gated Recurrent Neural Network Incorporating Spatial Features for Audio Tagging

Cited by in corpus (7)

• Audio Source Separation Using Variational Autoencoders and Weak Class Supervision
• End-to-End Waveform Utterance Enhancement for Direct Evaluation Metrics Optimization by Fully Convolutional Neural Networks
• End-to-end Networks for Supervised Single-channel Speech Separation
• Multi-Resolution Fully Convolutional Neural Networks for Monaural Audio Source Separation
• Neural Network Alternatives to Convolutive Audio Models for Source Separation
• Separation of Instrument Sounds using Non-negative Matrix Factorization with Spectral Envelope Constraints
• Star DGT: a Robust Gabor Transform for Speech Denoising