2019/04/15 by Rob Brekelmans, Brekelmans, Rob, Daniel Moyer +5 · 2 citations
Computer Science · Mathematics · #Blind Source Separation Techniques #FOS: Computer and information sciences #Image and Signal Denoising Methods #Information Theory (cs.IT) #Machine Learning (cs.LG) #Machine Learning (stat.ML) #Speech and Audio Processing #cs.IT #cs.LG #math.IT #stat.ML
paper · pdf · doi:10.48550/arxiv.1904.07199
NeurIPS 2019; updated Gaussian baseline results, added disentanglement
openalex publication_date 2019/04/15 · arxiv created 2019/11/14 · arxiv updated 2019/11/15 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
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, 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 β-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.