Pattern Recognition System Design with Linear Encoding for Discrete Patterns

TL;DR

This paper explores the design of compressive pattern recognition systems using linear encoding techniques, establishing a duality with communication codes, and analyzing their performance with various coding strategies.

Contribution

It introduces a novel connection between recognition system design and communication coding, utilizing low-density and LDPC matrices for efficient compression and recognition.

Findings

Effective use of low-density matrices for pattern compression

Dual connection between recognition systems and communication codes

Design strategies for recognition under general noise models

Abstract

In this paper, designs and analyses of compressive recognition systems are discussed, and also a method of establishing a dual connection between designs of good communication codes and designs of recognition systems is presented. Pattern recognition systems based on compressed patterns and compressed sensor measurements can be designed using low-density matrices. We examine truncation encoding where a subset of the patterns and measurements are stored perfectly while the rest is discarded. We also examine the use of LDPC parity check matrices for compressing measurements and patterns. We show how more general ensembles of good linear codes can be used as the basis for pattern recognition system design, yielding system design strategies for more general noise models.