"Human-Machine Collaboration for Visual Scene Understanding"

Dr. Amit K. Roy-Chowdhury, Professor, ECE/CS, UC Riverside

June 6th (Monday), 3:00pm
Building 406, Room 216

Most computer vision methods work by having an initial, labor-intensive training phase, followed by an operational phase where the trained models are used. However, in many applications, all the data may not be available at the outset and the learned models need to be adapted continuously as new data becomes available. This requires continuous learning methods in visual scene understanding, where humans and the machine vision system are working together collaboratively. In this talk, we address some of the research issues that arise in such a framework. First, we focus on active learning methods and how they can be used for continuously updating learned models by interacting with a human operator. We show results in object, activity and scene recognition and show that results similar to methods with a priori trained models can be obtained with significantly less manual labeling. Second, we consider the problem of summarizing large volumes of video data to present to the human in a manner that allows efficient decision making without being overwhelmed by the size of the data.

About Dr. Amit K. Roy-Chowdhury:

photo of Amit K.Roy-ChowdhuryAmit K. Roy-Chowdhury received Masters degree in systems science and automation from the Indian Institute of Science, Bangalore, India, and the Ph.D. degree in electrical engineering from the University of Maryland, College Park. He is a Professor of Electrical Engineering and a Cooperating Faculty in the Department of Computer Science at the University of California, Riverside. His broad research interests include the areas of image processing and analysis, computer vision, and statistical signal processing and pattern recognition. Together with his students and collaborators, he has over 100 technical publications in these areas. He is the first author of the book - Camera Networks: The Acquisition and Analysis of Videos over Wide Areas - the first research monograph on this topic. He has been on the organizing and program committees of multiple computer vision and image processing conferences and is serving on the editorial boards of multiple journals in related areas.

Hosted by: Dr. Yasamin Mostofi