Abhinav Shrivastava, a Ph.D. student in robotics, is one of a dozen computer science, electrical engineering and mathematics students nationwide to receive a 2014 Microsoft Research Ph.D. Fellowship.
Shrivastava’s research interest is computer vision. Specifically, he is developing visual learning algorithms that can leverage existing databases of labeled images, vast numbers of unlabeled images from the Internet and the implicit structure of the world — cars are generally found on roads, the sky is above the ground, buildings are upright — to improve computer vision systems at large scales. He is part of the project team for the Never Ending Image Learner (NEIL), a computer program that constantly searches the Web for images and attempts to identify objects and characterize scenes on its own.
The two-year fellowship covers tuition and fees, and includes a travel allowance, the offer of a paid internship, and an annual stipend.