Title: Multimedia Analytics for Videos
Date and Time: 30th November, 3:30 pm
Venue: KD 101, CSE Department

Abstract
In this talk, I will discuss multiple projects that are happening at the Multimedia Analytics group at Xerox Research Centre India.
 
First, I will talk about a distance learning method in non-vector spaces, where the triangle inequality is used to propagate the pairwise constraints to the unsupervised image pairs. This approach can work with any pairwise distance and does not require any vector representation of images. 
 
Next, I will talk about a novel approach to jointly segment and classify egocentric/first-person activity videos of daily-life. First, ego-centric activity classifiers are learnt in a novel multiple instance learning (MIL) based framework, which can remove distractors present in long and complex egocentric-activities. Second, these classifiers are used in a dynamic programming framework to simultaneously segment an egocentric video into individual activities and classify them.
 
Finally, I will discuss our recent work on deep learning for multi-faceted index classification and engagement analysis in educational videos. 
 
Bio:
 
Arijit Biswas is a research scientist at Xerox Research Centre India (XRCI). He works in the multimedia analytics group where his research interests are mainly in computer vision, machine learning, deep learning and text analytics. He received his PhD in Computer Science from University of Maryland, College Park in April 2014. His PhD thesis was on Semi-supervised and Active Learning Methods for Image Clustering. His thesis advisor was David Jacobs and he closely collaborated with Devi Parikh and Peter Belhumeur during his stay at UMD. While doing his PhD, Arijit also did internships at Xerox PARC (PARC-west) and Toyota Technological Institute at Chicago (TTIC). He has published papers in CVPR, ECCV, ACM-MM, BMVC, IJCV and CVIU.  Arijit has a Bachelor's degree in Electronics and Telecommunication Engineering from Jadavpur University, Kolkata.