96

CASS: Cross Adversarial Source Separation via Autoencoder

Abstract

This paper introduces a cross adversarial source separation (CASS) framework via autoencoder, a new model that aims at separating an input signal consisting of a mixture of multiple components into individual components defined via adversarial learning and autoencoder fitting. CASS unifies popular generative networks like auto-encoders (AEs) and generative adversarial networks (GANs) in a single framework. The basic building block that filters the input signal and reconstructs the ii-th target component is a pair of deep neural networks ENi\mathcal{EN}_i and DEi\mathcal{DE}_i as an encoder for dimension reduction and a decoder for component reconstruction, respectively. The decoder DEi\mathcal{DE}_i as a generator is enhanced by a discriminator network Di\mathcal{D}_i that favors signal structures of the ii-th component in the ii-th given dataset as guidance through adversarial learning. In contrast with existing practices in AEs which trains each Auto-Encoder independently, or in GANs that share the same generator, we introduce cross adversarial training that emphasizes adversarial relation between any arbitrary network pairs (DEi,Dj)(\mathcal{DE}_i,\mathcal{D}_j), achieving state-of-the-art performance especially when target components share similar data structures.

View on arXiv
Comments on this paper

We use cookies and other tracking technologies to improve your browsing experience on our website, to show you personalized content and targeted ads, to analyze our website traffic, and to understand where our visitors are coming from. See our policy.