Fashion image retrieval task aims to search relevant clothing items of a query image from the gallery. The previous recipes focus on designing different distance-based loss functions, pulling relevant pairs to be close and pushing irrelevant images apart. However, these methods ignore fine-grained features (e.g. neckband, cuff) of clothing images. In this paper, we propose a novel fashion image retrieval method leveraging both global and fine-grained features, dubbed Multi-Granular Alignment (MGA). Specifically, we design a Fine-Granular Aggregator(FGA) to capture and aggregate detailed patterns. Then we propose Attention-based Token Alignment (ATA) to align image features at the multi-granular level in a coarse-to-fine manner. To prove the effectiveness of our proposed method, we conduct experiments on two sub-tasks (In-Shop & Consumer2Shop) of the public fashion datasets DeepFashion. The experimental results show that our MGA outperforms the state-of-the-art methods by 1.8% and 0.6% in the two sub-tasks on the R@1 metric, respectively.
@article{arxiv.2302.08902,
title = {Fashion Image Retrieval with Multi-Granular Alignment},
author = {Jinkuan Zhu and Hao Huang and Qiao Deng and Xiyao Li},
journal= {arXiv preprint arXiv:2302.08902},
year = {2023}
}