On Macroscopic Complexity and Perceptual Coding

TL;DR

This paper explores the theoretical limits of lossy data compression by defining the complexity of objects from a macroscopic perspective, emphasizing perceptual coding and its implications for inference and pattern recognition.

Contribution

It introduces a framework for understanding complexity based on perceptual codes, linking macrostate complexity to lossy compression limits and observer-dependent perception.

Findings

Defines macroscopic complexity as perceptual code size

Highlights the advantage of macrostate-based inference

Suggests observer-dependent complexity impacts lossy compression

Abstract

The theoretical limits of 'lossy' data compression algorithms are considered. The complexity of an object as seen by a macroscopic observer is the size of the perceptual code which discards all information that can be lost without altering the perception of the specified observer. The complexity of this macroscopically observed state is the simplest description of any microstate comprising that macrostate. Inference and pattern recognition based on macrostate rather than microstate complexities will take advantage of the complexity of the macroscopic observer to ignore irrelevant noise.