Data Compression with Relative Entropy Coding

TL;DR

This thesis explores relative entropy coding, a flexible framework for data compression that extends classical methods to continuous spaces, enabling privacy-preserving and perceptually-aware compression in machine learning applications.

Contribution

It provides new theoretical limits, develops optimal algorithms using Poisson point processes, and demonstrates practical effectiveness on diverse data types with neural network implementations.

Findings

Achieves tight fundamental limits for relative entropy coding.

Develops optimal algorithms based on Poisson point processes.

Shows strong practical performance on image, audio, video, and protein data.

Abstract

Over the last few years, machine learning unlocked previously infeasible features for compression, such as providing guarantees for users' privacy or tailoring compression to specific data statistics (e.g., satellite images or audio recordings of animals) or users' audiovisual perception. This, in turn, has led to an explosion of theoretical investigations and insights that aim to develop new fundamental theories, methods and algorithms better suited for machine learning-based compressors. In this thesis, I contribute to this trend by investigating relative entropy coding, a mathematical framework that generalises classical source coding theory. Concretely, relative entropy coding deals with the efficient communication of uncertain or randomised information. One of its key advantages is that it extends compression methods to continuous spaces and can thus be integrated more seamlessly into modern machine learning pipelines than classical quantisation-based approaches. Furthermore, it is a natural foundation for developing advanced compression methods that are privacy-preserving or account for the perceptual quality of the reconstructed data. The thesis considers relative entropy coding at three conceptual levels: After introducing the basics of the framework, (1) I prove results that provide new, maximally tight fundamental limits to the communication and computational efficiency of relative entropy coding; (2) I use the theory of Poisson point processes to develop and analyse new relative entropy coding algorithms, whose performance attains the theoretic optima and (3) I showcase the strong practical performance of relative entropy coding by applying it to image, audio, video and protein data compression using small, energy-efficient, probabilistic neural networks called Bayesian implicit neural representations.