Productify - Mobile Visual Apparel Product Recognition

About

Productify is a mobile visual recognition app for apparel products employing state-of-the-art cloud-based computer vision service. Using Productify, you can visually interact with objects around you and turn them into purchases simply by pointing at them with a smartphone. Productify allows you to snap pictures of objects, recognize a match and receive suggestions of similar products. Discover where to buy, see hot trends, create wishlists, earn badges, share and recommend products to your friends.

The Productify app uses an API to communicate with the back-end service providing stateof-the-art computer vision algorithms for large-scale image search. Once the Productify back-end receives a request, it performs automatic foreground/background segmentation, creates a visual fingerprint of the query image and compares it to the fingerprints of the e-commerce products stored in the database. Next it returns a ranked list of similar products to the Productify front-end for the user to explore. The core of this service is the algorithms used for finding visual fingerprints based on product metadata, color, shape and texture as well as the consequent large scale search and matching methods. These algorithms have been tailored to the type of products for which visual appearance is the primary attribute (e.g. apparel).

Productify has been developed by the research and development team at OculusAI Technologies and has recently been acquired by Meltwater. The Productify research team consisted of:

  • Dr. Andrzej Pronobis (Head of Research and Chief Scientist of OculusAI)
  • Dr. Babak Rasolzadeh (CTO of OculusAI)
  • Dr. Bhaskar Chakraborty
  • Dr. Sahar Asadi
  • Dr. Scott Helmer

Related Publications

  • User Interface of a Mobile Camera Device for Viewpoint Biasing (A. Pronobis, B. Rasolzadeh), 2012. (Swedish Patent Application 1200716-7, Filed November 2012, Patent Pending)
  • Productify - Product Recognition in the Wild (G. Manno, A. Pronobis, B. Rasolzadeh), In CVPR 2012 IEEE Vision Industry and Entrepreneur Workshop (VIEW), 2012.