Unsupervised learning of features and object boundaries from local prediction

TL;DR

This paper presents an unsupervised model that learns both image features and object boundaries simultaneously by predicting local spatial relationships, resulting in features similar to biological vision and effective segmentation without supervision.

Contribution

It introduces a novel pairwise Markov random field model with binary switches, enabling joint learning of features and boundaries from images using contrastive objectives.

Findings

Learned features include local averages, opponent colors, Gabor patterns

Connectivity inference aligns with human visual segmentation

Model performs well on BSDS500 without contour training

Abstract

A visual system has to learn both which features to extract from images and how to group locations into (proto-)objects. Those two aspects are usually dealt with separately, although predictability is discussed as a cue for both. To incorporate features and boundaries into the same model, we model a layer of feature maps with a pairwise Markov random field model in which each factor is paired with an additional binary variable, which switches the factor on or off. Using one of two contrastive learning objectives, we can learn both the features and the parameters of the Markov random field factors from images without further supervision signals. The features learned by shallow neural networks based on this loss are local averages, opponent colors, and Gabor-like stripe patterns. Furthermore, we can infer connectivity between locations by inferring the switch variables. Contours inferred from this connectivity perform quite well on the Berkeley segmentation database (BSDS500) without any training on contours. Thus, computing predictions across space aids both segmentation and feature learning, and models trained to optimize these predictions show similarities to the human visual system. We speculate that retinotopic visual cortex might implement such predictions over space through lateral connections.