Single-Item Fashion Recommender: Towards Cross-Domain Recommendations

TL;DR

This paper introduces a content-based fashion recommender system that uses a neural network to recommend similar items from a single shop image, enhances personalization, improves robustness to out-of-domain queries, and proposes a new evaluation metric.

Contribution

It presents a novel single-item, cross-domain fashion recommendation approach with background augmentation and a customizable human score evaluation metric.

Findings

Effective in recommending similar fashion items from a single image

Enhanced robustness to street-to-shop out-of-domain queries

Introduced a new, interpretable evaluation metric for recommendations

Abstract

Nowadays, recommender systems and search engines play an integral role in fashion e-commerce. Still, many challenges lie ahead, and this study tries to tackle some. This article first suggests a content-based fashion recommender system that uses a parallel neural network to take a single fashion item shop image as input and make in-shop recommendations by listing similar items available in the store. Next, the same structure is enhanced to personalize the results based on user preferences. This work then introduces a background augmentation technique that makes the system more robust to out-of-domain queries, enabling it to make street-to-shop recommendations using only a training set of catalog shop images. Moreover, the last contribution of this paper is a new evaluation metric for recommendation tasks called objective-guided human score. This method is an entirely customizable framework that produces interpretable, comparable scores from subjective evaluations of human scorers.