CURRICULUM VITAE

Personal Profile

Name		: Nipun Kwatra
Email Address		: nipun@cc.gatech.edu
Home Page		: http://www.cc.gatech.edu/$\sim$nipun
Address		: GVU Center, College of Computing; 85, 5th Street NW; Atlanta, GA 30332-0760, USA

Research Interests

Computer Graphics, Computer Animation, Computer Vision, Machine Learning. Physically based and Data driven Animation and Rendering.

Education

08/2004 - Present	M.S. in Computer Science
	Georgia Institute of Technology, Atlanta, GA
	GPA 4.0/4.0
07/2000 - 05/2004	B.Tech in Computer Science and Engineering
	Indian Institute of Technology, Delhi, India
	GPA 8.93/10

Work Experience

Publications

Academic Awards

• Represented India in the 31$^{st}$International Physics Olympiad held in Leicester, England and won a Bronze Medal out of the 300 participants of 65 countries. The Indian Team stood an overall 3$^{rd}$.
• Won a Gold Medal for being among the top 5 out of 20,000 participants for the National Physics Olympiad. The Training of the top 30 was held in HBCSE, Mumbai.
• Selected among the top 30 students in India for the Training camp of the 32$^{nd}$ International Chemistry Olympiad.
• Received the INRIA-IIT scholarship in the year 2003.
• Received the Institute Merit Award for being among the top 7% in the institute in Fall 2000.
• Received the Institute Merit Award for being among the top 3 in the Computer Science and Engineering Department successively in Fall 2001, Fall 2002, spring 2003, Fall 2003
• Secured 3$^{rd}$ rank in Physica (physics competition), held in Tryst(All India Technical Festival, IIT Delhi).
• Secured 4th rank in Techkriti'2002(All India Technical Festival, IIT Kanpur) in the working model contest, Endeavor.
• Secured All India Rank 56 in the Joint Entrance Examination 2000 for admission to Indian Institutes Of Technology among the 150,000 aspirants.
• Secured All India Rank 1 in the CET 2000 for entrance to Panjab Engineering College.

Projects

• Interacting Motion Capture Data With Fluids
  Collaborators: Chris Wojtan, Prof. Irfan Essa, Prof. Greg Turk, Prof. Peter J. Mucha & Dr. Mark Carlson
  We present an algorithm for creating realistic animations of characters that are interacting with fluids. We approach the problem by combining dynamic simulation and motion capture data. To dynamically follow the motion capture data, we use the proportional derivative servo tracking controller or a novel constraint-based solver. The interaction of the articulated body with fluids is done by a simple extension of the Rigid Fluid method to incorporate joint constraints. Our constraint based solver for following the motion capture data takes as input the current state of the simulation and calculates the exact angular and linear accelerations of the bodies, which would enable the simulation to conform to the motion capture data exactly, two time steps later. These accelerations are then used to estimate the forces and torques that need to be applied at each joint. We demonstrate results for tasks like swimming and walking in water. The ability to have articulated bodies interact with fluids also allows us to generate simulations of simple water creatures, that are driven by simple controllers.
• Texture Optimization for Example-based Synthesis
  Collaborators: Dr. Vivek Kwatra, Prof. Irfan Essa & Prof. Aaron Bobick
  We present a novel technique for texture synthesis using optimization. We define a Markov Random Field (MRF)-based similarity metric for measuring the quality of synthesized texture with respect to a given input sample. This allows us to formulate the synthesis problem as minimization of an energy function, which is optimized using an Expectation Maximization (EM)-like algorithm. In contrast to most example-based techniques that do region-growing, ours is a joint optimization approach that progressively refines the entire texture. Additionally, our approach is ideally suited to allow for controllable synthesis of textures. Specifically, we demonstrate controllability by animating image textures using flow fields. We allow for general two-dimensional flow fields that may dynamically change over time. Applications of this technique include dynamic texturing of fluid animations and texture-based flow visualization.
  http://www.cc.gatech.edu/$\sim$nipun/publications/TO/TO-final.pdf
• `Where was this taken?' - Semi-supervised kernel based learning for location recognition
  Advisor : Bill Triggs, Movi Project, INRIA Grenoble, France.
  In this report we compare the performance of various supervised and semi-supervised learning techniques on the problem of classifying incoming images according to their indoor or outdoor location, given a small set of training images taken at each location. The results show that SVMs outperform all other inductive techniques. The semi-supervised transductive SVMs perform better than inductive SVMs for large enough data, but the performance falls when the training data is reduced to very low values
  http://www.cc.gatech.edu/$\sim$nipun/Projects/reportINRIA/report.pdf
• Activity Recognition
  Advisors: Prof. Subhashis Bannerjee & Prof. Prem Kalra (Deptt. of Computer Science & Engineering, IIT Delhi)
  This project is related to high level monitoring of activities going on in a scene. We proposed a simple framework for activity recognition based on a model of multi-layered finite state machines, built on top of a low level image processing module for spatio-temporal detections and limited object identification. The finite state machine network learns, in an unsupervised mode, usual patterns of activities in a scene over long periods of time. Then, in the recognition phase, usual activities are accepted as normal and deviant activity patterns are flagged as abnormal.
  We also extended this to recognize predefined activities and provided a programmable finite state machines for the same. We successfully demonstrated the technique for activities such as unidentified object detection and theft detection.
  http://vglab.cse.iitd.ernet.in/research/research_group2/1/activmon/index.shtml
• Augmented/Virtual Reality (Mini Project)
  Advisors : Prof. Subhashis Bannerjee & Prof. Prem Kalra
  We presented a novel approach to augment destination environments with objects cut from source environments, rather than augmenting 3-d models, which allows us to create virtual environments augmented with real objects such that the user cannot tell the difference between the real world and the virtual augmentation of it.
  The current work models the source objects as planes. A modified version of the ray tracing algorithm was developed which renders the destination environment using these planar objects and also the possible occluding planes in the destination environment. These planes in the destination environment were modeled using single view 3d reconstruction methods.
  The present results are available on:
  http://www.cc.gatech.edu/$\sim$nipun/aug_reality/AR.html
• Independent Study On Quantum Computing.
  Advisor : Prof. Sandeep Sen, IITD.
  In this Independent Study I went through the various basics and fundamentals of Quantum Computing. I studied various advanced algorithms like Quantum Fourier Transform, Grover Search Algorithm, Shor's Factorization Algorithm.
  http://www.cc.gatech.edu/$\sim$nipun/Projects/quantum/quantum_computing.pdf
• Recovering illumination distribution from shadows.
  Advisor : Prof. Subhashis Bannerjee
  In this project I studied and implemented the estimation of illumination distribution of a real scene from radiance distribution inside shadows cast by an object in the scene.
• Automated Navigation of a car using Computer Vision
  Advisor : Prof. Subhashis Bannerjee, IITD, India
  The project involved automatically navigating a car through a 2-D maze(through a feedback mechanism) avoiding obstacles by viewing from an overhead camera(fixed). It involved developing a finite state machine for the control mechanism of the car and vision techniques like perspective correction, color segmentation, K-L transformations etc.
  http://www.cc.gatech.edu./$\sim$nipun/toy_car/toycar.html
  Awarded 4th position(Among 50 entries) in All India Inter-college Technical Fest held at Indian Institute of Technology, Kanpur in March'2002
• Rapid Systems Programming
  Advisor : Prof. S. Arun Kumar, IITD, India
  This project involved developing a Generic System that takes as input the structural operational semantics(SOS) of a programming language and a program string obeying the specified semantics and returns the evaluated string. It assumes some data-structures such as Environment and Store and some predefined hardware level functions like ADD,LESS THAN etc.

Course Projects

• Photon Mapping
  Advisor: Prof. Greg Turk
  Course: Advanced Image Synthesis
  The aim of this project was to use the Photon Mapping technique proposed by Henrik Wann Jensen for global illumination. I successfully implemented the paper ``Global Illumination using Photon Maps'' by Henrik Wann Jensen. Results can be seen at:
  http://www.cc.gatech.edu/$\sim$nipun/Projects/photonMapping/pm.html

• Image Matting Using Variational Methods
  Advisor: Prof. Antony Yezzi
  Course: PDEs and their application in Computer Vision and Image Processing
  We used a variational approach to solve the image matting problem. We developed a novel energy function for describing the quality of a matte, foreground and background. The minimization of this energy function was done using a simple gradient descent method, which gave encouraging results.

Software and Hardware Skills and Experience

• Language Skills : C, C++, Perl, MATLAB, SML, JAVA, VRML, MIPS Assembly, Pascal, Prolog
• OS Skills : Linux, Unix, Win9x/NT, Sun Solaris, IRIX
• Packages used/familiar : OpenGL, OpenCV, Horatio, Gandalf
• H/W tools used : Xilinx, VHDL

Extra-Curricular Activities

• Won the best fresher award for Dance & Dramatics for the year 2000 in IIT Delhi.
• Won the Second Best Actor award in the inter-hostel Hindi Play competition held in 2002 in IIT Delhi.
• Coordinator (core-competitions) in Tryst(Technical Festival IIT-Delhi), 2004.(www.iitd.ernet.in/$\sim$tryst) which involved managing a team of approximately 200 people.
• Won 2nd prize in the inter-hostel swimming relay competition held in 2000 in IIT Delhi.
• Held the post of Representative of ``Photographic and Audio Visual Development Society" of IIT Delhi for 2001-2002.

References

Prof. Irfan A. Essa (Advisor)
Associate Professor, College of Computing
Georgia Institute of Technology
irfan@cc.gatech.edu
+1 (404) 894-6856 Tel    +1 (404) 894-0673 Fax

Prof. Greg Turk (Co-advisor)
Associate Professor, College of Computing
Georgia Institute of Technology
turk@cc.gatech.edu
+1 (404) 894-7508 Tel    +1 (404) 894-0673 Fax

Prof. Peter J. Mucha (Co-advisor)
Assistant Professor, Applied Mathematics Program
Institute for Advanced Materials
The University of North Carolina at Chapel Hill
mucha@unc.edu
+1 (919) 843-2550 Tel    +1 (919) 962-9345 Fax

Prof. Anthony J. Yezzi (Course Instructor)
Associate Professor, School of Electrical and Computer Engineering
Georgia Institute of Technology
ayezzi@ece.gatech.edu
+1 (404) 385-1017 Tel    +1 (404) 894-4641 Fax