@ IBM Research

With Watson, we are now able to accurately recognize, localize, describe and answer questions about images and video. The cloud-based IBM Watson Visual Recognition service uses visual learning and recognition to surface insights that address such specific questions as: “How do I repair a flat tire?” – explaining to a person or robot how to do a complex task from video; or “What happened at the event?” – summarizing real-world activities from multiple visual data sources.

IBM Research at CVPR 2017: Helping AI systems to 'see' with latest computer vision innovations

Read featured papers at CVPR 2017

Watson Visual Recognition API

Visual Recognition understands the contents of images - visual concepts tag the image, recognize food, find human faces, approximate age and gender, and find similar images in a collection. You can also train the service by creating your own custom concepts. Use Visual Recognition to detect a dress type in retail, identify spoiled fruit in inventory, and more.

Learn more



Watson Visual Recognition API demo


Advancing Visual Recognition and Unlocking Data in Plain Sight


Cognitive assistant for the visually impaired


Build your own programmable cardboard robot

Explore career opportunities @ IBM Research