In this paper, we propose learning an embedding function for content-based image retrieval within the e-commerce domain using the triplet loss and an online sampling method that constructs triplets from within a minibatch. We compare our method to several strong baselines as well as recent works on the DeepFashion and Stanford Online Product datasets. Our approach significantly outperforms the state-of-the-art on the DeepFashion dataset. With a modification to favor sampling minibatches from a single product category, the same approach demonstrates competitive results when compared to the state-of-the-art for the Stanford Online Products dataset.
@article{arxiv.1810.04652,
title = {Learning Embeddings for Product Visual Search with Triplet Loss and Online Sampling},
author = {Eric Dodds and Huy Nguyen and Simao Herdade and Jack Culpepper and Andrew Kae and Pierre Garrigues},
journal= {arXiv preprint arXiv:1810.04652},
year = {2018}
}