# Visual Search at eBay

## TL;DR

This paper presents a scalable, deep learning-based visual search system for eBay that balances accuracy and latency, demonstrating superior performance over baselines and real-world deployment in eBay's infrastructure.

## Contribution

The paper introduces a novel end-to-end visual search architecture optimized for large-scale, volatile inventories using supervised deep learning and binary signatures, enabling efficient and accurate search.

## Key findings

- Faster and more accurate than unsupervised baselines on ImageNet
- Deployed in eBay's cloud infrastructure for real-world applications
- Balances search relevance with latency through optimized system architecture

## Abstract

In this paper, we propose a novel end-to-end approach for scalable visual search infrastructure. We discuss the challenges we faced for a massive volatile inventory like at eBay and present our solution to overcome those. We harness the availability of large image collection of eBay listings and state-of-the-art deep learning techniques to perform visual search at scale. Supervised approach for optimized search limited to top predicted categories and also for compact binary signature are key to scale up without compromising accuracy and precision. Both use a common deep neural network requiring only a single forward inference. The system architecture is presented with in-depth discussions of its basic components and optimizations for a trade-off between search relevance and latency. This solution is currently deployed in a distributed cloud infrastructure and fuels visual search in eBay ShopBot and Close5. We show benchmark on ImageNet dataset on which our approach is faster and more accurate than several unsupervised baselines. We share our learnings with the hope that visual search becomes a first class citizen for all large scale search engines rather than an afterthought.