POOJA NATH


Contact Information
Email: pooja@cs.umd.edu
Address: 3412 Tulane Dr, #21, Hyattsville, MD 20783-1828
Telephone: (301) 613-1498



Objective
To work in a challenging environment in the field of s/w development



Education
M.S., Computer Science, University of Maryland, College Park
(Graduation: May 2004)
	- Specializing in Computer Vision


B.Tech., Computer Science and Engg., Indian Institute of Technology (IIT) , Kanpur
(1998 - 2002)
	- Deans Merit certificate of Academic Excellence at IIT Kanpur, 1999




Experience

*Graduate Teaching Assistant, Dept. of Computer Science, University of Maryland
(2003-2004)
	-Design and Analysis of Algorithms
	-Computer Graphics


*Graduate Research Assistant, Computer Vision Group, University of Maryland
(2002-2003)
Worked on the Human Identification (HID) Project. Analyzed cues from human gait, 
which tend to be individual-specific, for human recognition. Extracted features 
from human shape across multiple scales to perform action recognition.


*Summer intern, Computer Vision Group, University of California, Berkeley
(Summer 2003)
Worked with a group of 3 people on action recognition. Information from motion
 was used to determine specific running or walking actions in a football video.  
My work is an acknowledged contribution in the publication "Recognizing Action 
at a Distance", by Efros et. al., presented at ICCV '03, Nice, France. 


*Summer intern, Signal and Image Processing Group, Paris Telecom, Paris
(Summer 2001)
Developed an application for the reconstruction of images from geophysical data 
sets. Extended this work at IIT Kanpur by using Genetic Algorithms for robust 
parameter estimation.



Technical skills
Languages: C, C++, Java, Perl, Lisp
Experience in: major operating systems, Matlab, OpenGL
Additional strengths: 
	- Strong mathematical and analytical skills. 
	- Excellent communication and presentation abilities.



Academic Projects
Vision:
* Optimization techniques for Background Modeling of image sequences
* 3D Boids simulation with flocking and obstacle avoidance behavior 
* Developed a 3D graphics game for rendering player actions in a real-life setting
* Hindi Text Extraction from images using Support Vector Machines (SVMs)
* Application for Handwriting Recognition and Beautification

Other areas:
* Developed a Natural Language Processing based system to examine conformities 
in large sentence data sets
* Designed and implemented a robot motion planning algorithm using probabilistic 
roadmaps
* Developed a School Complaint Management System using Rational Rose
* Designed and developed a hotel management system on Oracle
* Team of 3: Implemented searching and indexing of files in Chord (a scalable 
peer-to-peer lookup service)
* Evaluated various high bandwidth applications over Bluetooth
* Team of 4: Implemented a Java Interpreter with the help of a bytecode library



References
1. Larry Davis, Professor and Chair, Department of Computer Science, 
University of Maryland, College Park
2. David Mount, Professor, Department of Computer Science, 
University of Maryland, College Park