Web-Scale Responsive Visual Search at Bing
This work addresses the problem of scalable and fast visual search for users of Bing, though it appears incremental as it builds on existing deep learning features and distributed platforms.
The paper tackles the challenge of building a web-scale visual search system for Microsoft Bing that handles tens of billions of images with low latency, achieving response times under 200 ms using a cascaded learning-to-rank framework.
In this paper, we introduce a web-scale general visual search system deployed in Microsoft Bing. The system accommodates tens of billions of images in the index, with thousands of features for each image, and can respond in less than 200 ms. In order to overcome the challenges in relevance, latency, and scalability in such large scale of data, we employ a cascaded learning-to-rank framework based on various latest deep learning visual features, and deploy in a distributed heterogeneous computing platform. Quantitative and qualitative experiments show that our system is able to support various applications on Bing website and apps.