Oct. 6th, 3pm, KL 232

Title: (Computer) Vision Without Sight

Roberto Manduchi
University of California, Santa Cruz

ABSTRACT

Computer vision, the art of making sense of images and videos captured by a camera, has made substantial progress over the past two decades. Mobile vision apps specifically designed for people with low vision or blindness are becoming commonplace, leveraging the popularity of smartphones (especially iPhones) in these communities. Using a camera without the ability to preview images in the viewfinder, though, can be exceedingly difficult, which limits the utility of these systems. I will present two different user studies with blind participants using different computer vision systems integrated in an iPhone, and equipped with an interface mechanism to facilitate correct camera aiming. These experiments highlighted a number of interesting, at times unexpected, aspects of the complex interaction between system design, interface mechanism, and the user’s own prioprioceptive abilities.

Part of this work was conducted in collaboration with Dr. James Coughlan of SKERI.

BIO:

Roberto Manduchi is a Professor of Computer Engineering at the University of California, Santa Cruz. Prior to joining UCSC in 2001, he held positions at Apple and at the Jet Propulsion Laboratory. His research and educational interests are in computer vision and sensor processing, with applications to assistive technology for people with visual impairment. He holds a Dottorato di Ricerca degree from the University of Padova, Italy.