Progressive Fashion Attribute Extraction

TL;DR

This paper introduces a progressive training method for extracting multiple fashion attributes from images, allowing models to handle missing labels and improve accuracy over traditional multi-label approaches.

Contribution

It proposes a novel progressive training approach that fine-tunes attribute-specific branches from a base network, effectively managing missing and ambiguous labels in fashion attribute extraction.

Findings

Improved classification accuracy over multi-label methods.

Effective handling of missing and ambiguous labels.

Demonstrated benefits across various fashion attributes.

Abstract

Extracting fashion attributes from images of people wearing clothing/fashion accessories is a very hard multi-class classification problem. Most often, even catalogues of fashion do not have all the fine-grained attributes tagged due to prohibitive cost of annotation. Using images of fashion articles, running multi-class attribute extraction with a single model for all kinds of attributes (neck design detailing, sleeves detailing, etc) requires classifiers that are robust to missing and ambiguously labelled data. In this work, we propose a progressive training approach for such multi-class classification, where weights learnt from an attribute are fine tuned for another attribute of the same fashion article (say, dresses). We branch networks for each attributes from a base network progressively during training. While it may have many labels, an image doesn't need to have all possible labels for fashion articles present in it. We also compare our approach to multi-label classification, and demonstrate improvements over overall classification accuracies using our approach.