Challenges in Disentangling Independent Factors of Variation

TL;DR

This paper addresses the challenge of disentangling independent factors of variation in images using weak labels, proposing an autoencoder-based model that transfers attributes but faces issues like reference ambiguity.

Contribution

It introduces a novel autoencoder framework trained with weak labels to disentangle factors of variation and analyzes the impact of feature dimensionality and adversarial training.

Findings

Model successfully transfers attributes across datasets.

Reference ambiguity can occur without additional knowledge.

Feature dimensionality influences disentanglement effectiveness.

Abstract

We study the problem of building models that disentangle independent factors of variation. Such models could be used to encode features that can efficiently be used for classification and to transfer attributes between different images in image synthesis. As data we use a weakly labeled training set. Our weak labels indicate what single factor has changed between two data samples, although the relative value of the change is unknown. This labeling is of particular interest as it may be readily available without annotation costs. To make use of weak labels we introduce an autoencoder model and train it through constraints on image pairs and triplets. We formally prove that without additional knowledge there is no guarantee that two images with the same factor of variation will be mapped to the same feature. We call this issue the reference ambiguity. Moreover, we show the role of the feature dimensionality and adversarial training. We demonstrate experimentally that the proposed model can successfully transfer attributes on several datasets, but show also cases when the reference ambiguity occurs.