SIR Similar Image Retrieval For Product Search In E-commerce

Stanley Theban, Vanjara Nihar, Pan Yanxin, Pirogova Ekaterina, Chakraborty Swagata, Chaudhuri Abon. Arxiv 2020

[Paper]    
ARXIV Image Retrieval Supervised

We present a similar image retrieval (SIR) platform that is used to quickly discover visually similar products in a catalog of millions. Given the size, diversity, and dynamism of our catalog, product search poses many challenges. It can be addressed by building supervised models to tagging product images with labels representing themes and later retrieving them by labels. This approach suffices for common and perennial themes like “white shirt” or “lifestyle image of TV”. It does not work for new themes such as “e-cigarettes”, hard-to-define ones such as “image with a promotional badge”, or the ones with short relevance span such as “Halloween costumes”. SIR is ideal for such cases because it allows us to search by an example, not a pre-defined theme. We describe the steps - embedding computation, encoding, and indexing - that power the approximate nearest neighbor search back-end. We also highlight two applications of SIR. The first one is related to the detection of products with various types of potentially objectionable themes. This application is run with a sense of urgency, hence the typical time frame to train and bootstrap a model is not permitted. Also, these themes are often short-lived based on current trends, hence spending resources to build a lasting model is not justified. The second application is a variant item detection system where SIR helps discover visual variants that are hard to find through text search. We analyze the performance of SIR in the context of these applications.

Similar Work