Events Calendar

CS Colloquium: Lorenzo Torresani (Dartmouth): Challenges and Opportunities in Visual Recognition with Big Image Data

Speaker: Lorenzo Torresani, Dartmouth

Talk Title: Challenges and Opportunities in Visual Recognition with Big Image Data

Series: CS Colloquium

Abstract: The last few years have seen a tremendous explosion of image and video data on the Web. Unfortunately only a small portion of this visual data is annotated with text. Even when tags are available, they often do not describe accurately the semantics of the image or the video. This renders traditional text-search an ineffective tool on these collections. In this talk I will describe some of my recent work on designing visual recognition systems that can help users browse and search image repositories more effectively.

I will begin with an algorithm that addresses the computational challenges posed by visual recognition in Web-scale image databases. Our approach centers around the learning of a compact image code optimized to yield accurate recognition with linear (i.e., efficient) classifiers: even when the representation is compressed to less than 300 bytes per image, linear classifiers trained on our descriptor yield accuracy matching the state-of-the-art but at orders of magnitude lower computational cost.

In the second part of my talk I will present a method that embraces Big Image Data as an opportunity to improve visual recognition. Our algorithm exploits a dataset of 10 million labeled photos to learn a universal semantic distance between images. This metric can be used either as a similarity measure to find pictures by example or as a “kernel” in distance-based image classifiers, yielding a significant boost in accuracy over traditional metrics.

Biography: Lorenzo Torresani is an Assistant Professor in the Computer Science Department at Dartmouth College. He received a Laurea Degree in Computer Science with summa cum laude honors from the University of Milan (Italy) in 1996, and an M.S. and a Ph.D. in Computer Science from Stanford University in 2001 and 2005, respectively. In the past, he has worked at several industrial research labs including Microsoft Research Cambridge, Like.com, and Digital Persona. His research interests are in computer vision and machine learning. In 2001, Torresani and his coauthors received the Best Student Paper Award at the IEEE Conference On Computer Vision and Pattern Recognition (CVPR). He is the recipient of a National Science Foundation CAREER Award.=

Host: Gerard Medioni

Location: Hughes Aircraft Electrical Engineering Center (EEB) - 248

Audiences: Everyone Is Invited

Contact: Assistant to CS chair