Research Area: Computer Vision (Person Re-identification and Tracking, Machine Learning etc.)

o In the first stage, developed an information theoretic and attribute feedback based continuous active person re-identification system. o In the second stage, developed a sparsity regularized non-redundant representative selection framework for multi-sensor person re-identification with a view to minimize the annotation effort of the human in the loop. Consistent Person Re-identification in a Camera Network. (Matlab, C++) Jan 2014 – Dec 2014 o Developed a Binary Integer Program based framework for boosting multi-camera person re-identification performance by enforcing consistency across all possible pairs of camera Person Re-Identification in the Function Space of Feature Warps. (Matlab, C++) Dec 2011 – Dec 2013 o Designing a person re-identification system across a non-overlapping camera network by studying th