Bringing Multimodality to Amazon Visual Search System

TL;DR

This paper enhances Amazon's visual search by integrating vision-language pretraining with image-text alignment losses into deep metric learning, significantly reducing false positives and improving multimodal search performance.

Contribution

It introduces a novel multimodal deep metric learning framework with image-text alignment, including 3-tower and 4-tower models, to improve image matching accuracy in visual search systems.

Findings

4.95% improvement in image matching click-through rate

Enhanced model reduces false positives in image-to-image matching

Both offline and online experiments confirm performance gains

Abstract

Image to image matching has been well studied in the computer vision community. Previous studies mainly focus on training a deep metric learning model matching visual patterns between the query image and gallery images. In this study, we show that pure image-to-image matching suffers from false positives caused by matching to local visual patterns. To alleviate this issue, we propose to leverage recent advances in vision-language pretraining research. Specifically, we introduce additional image-text alignment losses into deep metric learning, which serve as constraints to the image-to-image matching loss. With additional alignments between the text (e.g., product title) and image pairs, the model can learn concepts from both modalities explicitly, which avoids matching low-level visual features. We progressively develop two variants, a 3-tower and a 4-tower model, where the latter takes one more short text query input. Through extensive experiments, we show that this change leads to a substantial improvement to the image to image matching problem. We further leveraged this model for multimodal search, which takes both image and reformulation text queries to improve search quality. Both offline and online experiments show strong improvements on the main metrics. Specifically, we see 4.95% relative improvement on image matching click through rate with the 3-tower model and 1.13% further improvement from the 4-tower model.