I am currently working at Yahoo! Search Marketing.

I am a second-year Masters student (yeah! almost finishing...) at the Robotics Institute (RI), within the School of Computer Science (SCS), Carnegie Mellon University (CMU). My advisor is Prof. Martial Hebert.

In RI, I am associated with the 3D Computer Vision Group. I am (was) involved with E3D, PerceptOR and CTA Robotics projects. Most of my research revolves around 3D Computer Vision and Statistical Pattern Recognition.

Before coming to Carnegie Mellon, I got my Bachelor's degree from the Dept. of Computer Science & Engg., Indian Institute of Technology (IIT), Kharagpur, India.

Here is my official Robotics Institute webpage.

I am a member of the 3D Computer Vision Group, part of the Vision and Autonomous Systems Center (VASC). I am advised by Prof. Martial Hebert. My Masters thesis focusses on automatic classification and prototyping of 3D objects .

• Automatic 3D Object Classification and prototyping I am working on algorithms to automatically categorize a set of 3D objects (each object specified as a cloud of points in 3D) into disjoint classes. Given a categorization, the next step is to find representatives (prototypes) to each of the classes automatically. By storing only the class prototypes, one can minimize not only the time to search for a given query but also the storage space (required to store the model database), without sacrificing the accuracy of the system. This is part of the E3D program.

• Exploiting Aerial Range data for Off-road Perception I worked on algorithms to evaluate cost-maps (using overhead range data) of an outdoor environment, which are used in planning apriori mission paths for an All Terrain Vehicle (ATV). This assists the vehicle in perceiving farther than what is possible using only the on-board sensors. This assignment also involved methods to filter vegetation from the range data, to generate accurate cost maps. The paths planned were used in autonomous missions that were as long as a kilometer in rough, challenging off-road terrains.

• Classification of Range Data Worked on algorithms to classify range data (from a laser range finder, on board a mobile robot) into vegetation and non-vegetation classes.

• Automatic mosaicing of Aerial Reconnaissance Images Worked on algorithms to stitch a set of aerial images into a big mosaic. This is my undergraduate thesis. It was jointly awarded the best undergraduate thesis by the Department of Comupter Science & Engg, IIT, Kharagpur, India, May 2001.

• Experimental Results in Using Aerial Lidar Data for Mobile Robot Navigation
  N. Vandapel, RR Donamukkala and M. Hebert
  International Conference on Field and Service Robotics (FSR), 2003.
  

• Robotic Walker that Provides Guidance
  A Morris, RR Donamukkala , A. Kapuria, A. Steinfeld, J. Matthews, J. Jacob, S. thrun
  IEEE Conference on Robotics and Automation (ICRA), 2003.
  

• Evaluation and Comparison of Terrain Classification Techniques from LADAR Data for Autonomous Navigation
  M. Hebert, N. Vandapel, S. Keller, and R.R. Donamukkala
  23rd Army Science Conference, December, 2002. [Abstract]