Single channel audio source separation using convolutional denoising autoencoders
- 1 November 2017
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1265-1269
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.Keywords
All Related Versions
This publication has 23 references indexed in Scilit:
- A fully convolutional deep auditory model for musical chord recognition2016
- Very Deep Convolutional Neural Networks for Noise Robust Speech RecognitionIEEE/ACM Transactions on Audio, Speech, and Language Processing, 2016
- Multichannel Audio Source Separation With Deep Neural NetworksIEEE/ACM Transactions on Audio, Speech, and Language Processing, 2016
- Stacked Convolutional Denoising Auto-Encoders for Feature RepresentationIEEE Transactions on Cybernetics, 2016
- The 2015 Signal Separation Evaluation Campaign2015
- Adaptive Denoising Autoencoders: A Fine-Tuning Scheme to Learn from Test Mixtures2015
- Source separation using regularized NMF with MMSE estimates under GMM priors with online learning for the uncertaintiesDigital Signal Processing, 2014
- Stacked Convolutional Auto-Encoders for Hierarchical Feature Extraction2011
- Evaluation of Pooling Operations in Convolutional Architectures for Object Recognition2010
- Reducing the Dimensionality of Data with Neural NetworksScience, 2006