About

Publications

See Papers for a list of my publications and other research work in school.

Artificial Intelligence and Robotics

Till recently, I used to work at the Yale Social Robotics lab, which is part of the Computer Science department. My primary focus there was the problem of determining whether a given motion is animate or not, and my advisor was Prof. Brian Scassellati. This was part of the Nico project - the humanoid robot pictured below.

One of the most fundamental social skills which humans possess is the awareness of other minds, which allows the attribution of beliefs, goals and desires to other individuals. Any humanoid robot thus put in a socially dynamic setting amongst humans must learn to distinguish between various levels of social objects and characters, and accordingly interact with them. From the perspective of a robot, an object in its environment may be classified as Social or Non-Social - those with a controlling intelligence behind them, and others which are dependent on purely physical circumstances. The ability of animals to classify motion as belonging to either of the above classifications is an important aspect in their survival - for instance, split second decisions may need to be taken to determine a sudden motion as a possible predator. We refer to these classes of motion as animate and inanimate, although the distinction is not a clear one.

The objective of my work was to develop a system which can classify such motions for a plethora of different purposes.

3D Particle Simulation System for Condensed Matter Physics

While working for the O'Hern group, which is part of the Department of Mechanical Engineering, as well as Physics at Yale University, I developed a 3D simulation system to render particles based on the results of certain simulations, which were used for theoretical and computational studies of Jamming and Glass Transitions in Soft Matter.

Older Research

In the past, I have worked on a number of different research areas. I interned at the Indian Centre for Artificial Intelligence and Robotics, which is part of the Defence Research and Development Organization at the Indian Ministry of Defence. Working at the Virtual Reality lab there, I worked on developing techniques to enhance efficiency when simulating collisions between high resolution objects. It essentially involved looking at high polygon objects, and creating a technique which would automatically fit the object to the closest simple primitive, rather than have it generated manually. A lot of my simulations were done using the open source graphics engine, Crystalspace 3D with sparing use of VTK. This work has implications in a number of fields, ranging from medical imagery, virtual surgery, haptics and others. My advisors were Dr. J. Prakash Naidu, Head of the Automation and Virtual Reality group at CAIR, and Manohar B. Srikanth.

Before that, I worked at the National Informatics Centre in India, as part of the Cyber Security Group at the Indian Ministry of Information Technology. During my stint there, I worked on Intrusion Detection Systems and other aspects of network security.

I've also been actively involved in a number of open source projects in various capabilities. Googling my name will tell you more.