Visual Search at Pinterest

TL;DR

This paper shows how a small team can build and deploy a large-scale visual search system using cloud computing and open-source tools, and demonstrates its effectiveness in improving user engagement at Pinterest.

Contribution

It provides a practical implementation of a cost-effective, large-scale visual search system and shares insights from deploying it in a commercial setting.

Findings

Visual search improves user engagement at Pinterest.

Open-source tools enable scalable visual search deployment.

Cost-effective approach for small teams to implement visual search.

Abstract

We demonstrate that, with the availability of distributed computation platforms such as Amazon Web Services and open-source tools, it is possible for a small engineering team to build, launch and maintain a cost-effective, large-scale visual search system with widely available tools. We also demonstrate, through a comprehensive set of live experiments at Pinterest, that content recommendation powered by visual search improve user engagement. By sharing our implementation details and the experiences learned from launching a commercial visual search engines from scratch, we hope visual search are more widely incorporated into today's commercial applications.