Source Separation with Deep Generative Priors
arXiv:2002.07942
Abstract
Despite substantial progress in signal source separation, results for richly structured data continue to contain perceptible artifacts. In contrast, recent deep generative models can produce authentic samples in a variety of domains that are indistinguishable from samples of the data distribution. This paper introduces a Bayesian approach to source separation that uses generative models as priors over the components of a mixture of sources, and noise-annealed Langevin dynamics to sample from the posterior distribution of sources given a mixture. This decouples the source separation problem from generative modeling, enabling us to directly use cutting-edge generative models as priors. The method achieves state-of-the-art performance for MNIST digit separation. We introduce new methodology for evaluating separation quality on richer datasets, providing quantitative evaluation of separation results on CIFAR-10. We also provide qualitative results on LSUN.
20 pages; ICML camera-ready version
References in corpus (10)
- Auto-Encoding Variational Bayes
- GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium
- WaveNet: A Generative Model for Raw Audio
- How transferable are features in deep neural networks?
- Neural Discrete Representation Learning
- LSUN: Construction of a Large-scale Image Dataset using Deep Learning with Humans in the Loop
- Generative Modeling by Estimating Gradients of the Data Distribution
- Joint Optimization of Masks and Deep Recurrent Neural Networks for Monaural Source Separation
- Music Source Separation in the Waveform Domain
- Neural Network Alternatives to Convolutive Audio Models for Source Separation
Cited by in corpus (5)
- Wavelet Transform-assisted Adaptive Generative Modeling for Colorization
- Learning and controlling the source-filter representation of speech with a variational autoencoder
- Semi-Blind Source Separation with Learned Constraints
- Music Source Separation with Generative Flow
- Problems using deep generative models for probabilistic audio source separation