I work on computer vision and machine learning research problems with applications in robot perception & manipulation. My current research work revolves around data driven robot manipulation. I am particularly interested in the problem of transferring grasps across grippers and development of robust computer vision and geometry based algorithms for grasping.

Research
My main research focus is on robot manipulation methods, especially robust grasping. I am interested in developing generalizable grasp and manipulation methods which can scale well to unseen objects and gripper morphologies. I am currently trying to develop generative models for fast and accurate grasp synthesis. In addition to this, I am also interested in interactive perception based methods for real world robotics and have a couple of projects relating to real world segmentation. Apart from these, I am excited to learn more about topics like equivariant networks, deformable object modeling, deep generative models and geometry fundamentals!