Exploitation of Image Statistics with Sparse Coding in the Case of Stereo Vision

TL;DR

This paper demonstrates that sparse coding, combined with a simple classifier, can effectively infer stereo disparity from naturalistic images, highlighting the importance of redundant representations and active coefficients in the process.

Contribution

It shows that sparse coding with LCA can be used for stereo disparity inference, revealing the role of redundancy and active coefficients in robust pattern recognition.

Findings

Disparity inference was successful using sparse coding.

A highly redundant representation improves robustness.

Inference error correlates with the number of active coefficients.

Abstract

The sparse coding algorithm has served as a model for early processing in mammalian vision. It has been assumed that the brain uses sparse coding to exploit statistical properties of the sensory stream. We hypothesize that sparse coding discovers patterns from the data set, which can be used to estimate a set of stimulus parameters by simple readout. In this study, we chose a model of stereo vision to test our hypothesis. We used the Locally Competitive Algorithm (LCA), followed by a na\"ive Bayes classifier, to infer stereo disparity. From the results we report three observations. First, disparity inference was successful with this naturalistic processing pipeline. Second, an expanded, highly redundant representation is required to robustly identify the input patterns. Third, the inference error can be predicted from the number of active coefficients in the LCA representation. We conclude that sparse coding can generate a suitable general representation for subsequent inference tasks. Keywords: Sparse coding; Locally Competitive Algorithm (LCA); Efficient coding; Compact code; Probabilistic inference; Stereo vision