Exact Rate-Distortion in Autoencoders via Echo Noise

arXiv1904.07199MaRDI QIDQ6317229

Author name not available (Why is that?)

Publication date: 15 April 2019

Abstract: Compression is at the heart of effective representation learning. However, lossy compression is typically achieved through simple parametric models like Gaussian noise to preserve analytic tractability, and the limitations this imposes on learning are largely unexplored. Further, the Gaussian prior assumptions in models such as variational autoencoders (VAEs) provide only an upper bound on the compression rate in general. We introduce a new noise channel, emph{Echo noise}, that admits a simple, exact expression for mutual information for arbitrary input distributions. The noise is constructed in a data-driven fashion that does not require restrictive distributional assumptions. With its complex encoding mechanism and exact rate regularization, Echo leads to improved bounds on log-likelihood and dominates $$-VAEs across the achievable range of rate-distortion trade-offs. Further, we show that Echo noise can outperform flow-based methods without the need to train additional distributional transformations.

Has companion code repository: https://github.com/brekelma/echo

This page was built for publication: Exact Rate-Distortion in Autoencoders via Echo Noise

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6317229)