Dr. Aritra Dutta will be speaking at this seminar about a general introduction to his current research --- Connecting the Dots and Reading Between the Lines.

Abstract: My current research interests include (stochastic) optimization, distributed computing, numerical linear algebra, machine learning, low-rank approximation, and its applications in computer vision (image and video analysis). Additionally, I am interested in different video understanding tasks (object detection and tracking, action recognition, etc.) in computer vision. As researchers, working on research problems is similar to “standing close to elephants.” Often our initial reactions to events and situations are incorrect, or perhaps more accurately, they are incomplete. The analogy is that if we stand too close to the elephant, we cannot see the entire elephant. Moreover, there exist many significant discrepancies between theory and practice, and that opens many opportunities.

In the above spirit, in this talk, as the time permits, I will talk about some open problems in my present work, ranging from distributed computing to (continuous) optimization, to self-supervised video understanding to numerical linear algebra. Understanding these problems through the lens of statistics, mathematics, and computer science would be advantageous for creating foundational aspects of many future research directions.

]]>