My research interests are in the field of developing machine learning algorithms with applications to various fields in Computer Vision. Currently my focus is on distance metric learning algorithms which is crucial for performance of various algorithms that rely on choosing an appropriate distance measure. Algorithms used for the task of clustering and classification are two examples of it.

I intend to learn Mahalanobis distance metric parametrized by positive semidefinte matrix. This problem can be seen as an instance of Semi Definte Progrmaming but I follow a joint optimization approach on positive orthant and Stiefel Manifold. More details on the same can be found in my Diff-CV 2015 paper.

I am also exploring the fields of image segmentation, object detection and localization, and classification strategies using low level computer vision features such as SIFT, HOG etc.