SUNITA NAYAK

4085 Nobel Dr., Apt#31

Phone: (813) 541-4342

San Diego, CA 92122

Email: nayak.sunita@gmail.com ,

OBJECTIVE

To seek a full-time R&D position in one or more of the following fields: Image/Video Processing,

Computer Vision, Pattern Recognition and Machine Learning.

CITIZENSHIP: Indian

WORK AUTHORIZATION STATUS: Currently in Optional Practical Training period with a

valid Employment Authorization Document. Can start work immediately.

QUALIFICATION

• Strong background in Video Analysis, Image Processing, Computer Vision and Pattern

Recognition: theories, algorithms and implementation.

• Strong background in Data Structures and Computer Algorithms.

• Proficient in Matlab, C/C++ and Python.

• Experienced in Java Servlets, JavaScript, Oracle and PL/SQL

EDUCATION

University of South Florida

07/2005 - 01/2008

Ph.D. (Computer Science & Engineering)

Advisor: Prof. Sudeep Sarkar

Department of Computer Science & Engineering

Thesis: Representation and Learning for Automated Sign Language Recognition

Successfully defended on Jan 17

th

, 2008

GPA: 3.91/4.0

University of South Florida

08/2003 - 07/2005

M.S. (Computer Science)

Advisor: Prof. Sudeep Sarkar

Department of Computer Science & Engineering

Thesis: A Vision-Based Approach for Unsupervised Modeling of Signs Embedded in Continuous

Sentences

Regional Engineering College, Rourkela, India

09/1997 - 05/2001

(Now called National Institute of Technology, Rourkela)

B.E. (Computer Science & Engineering)

RESEARCH EXPERIENCE

Graduate Research Intern: Automated Molecular Imaging Group, The Scripps Research

Institute, La Jolla, CA

01/2007 - 11/2007

Molecular Imaging & Reconstruction

• Developed an unsupervised classification algorithm to improve 3D reconstructions of protein

macromolecules from extremely noisy digital images captured using an electron microscope.

Several digital images of the molecule that have the same viewing parameters are averaged to

improve the signal to noise ratio. We developed an automated algorithm to identify and reject

images of molecules having inconsistent viewing parameters, or having extremely low signal

to noise ratio.

• Developed an algorithm to estimate the astigmatism present in the micrographs captured

using an electron microscope, using a RANdom SAmple Consensus (RANSAC)-based

approach. When the astigmatism – a parameter in the contrast transfer function of an electron

microscope – is not estimated and corrected for, the resolution of 3D reconstruction suffers

significantly.

• Developed an expertise in pipeline-based automated processing of molecular imaging and

reconstruction. It includes automated contrast enhancement of micrographs, automated

particle picking and classification of particles, and 3D reconstruction of macromolecules.

Tools Used: Matlab, Python

Graduate Research Assistant: Computer Vision & Image Analysis Research Laboratory,

USF

08/2003 - 12/2006

Video Analysis

• Developed an algorithm for American Sign Language (ASL) recognition by analyzing video

sequences of an ASL signer. Unlike previous approaches, we do not require the signer to wear

specialized colored gloves, and we do not need explicit tracking of the body parts.

• Developed a random sampling-based algorithm for capturing the shapes and spatial

configurations of objects in an image using low-level primitives, e.g. edge points, and

efficiently computing the histograms used for representing motion.

• Proposed a novel technique for embedding histograms into a low-dimensional space by

preserving the probabilistic distances like Bhattacharya, Matusita, Chernoff, Kullback Leibler

(KL) and symmetric-KL distance measures.

• Implemented the interpolation of motion sequences and empirically showed that it improves

recognition accuracy.

Gesture Recognition and Human Activity Classification

• Developed a detection-based algorithm for gesture recognition and activity classification

without performing tracking, and by using global pose changes instead. Showed results of the

algorithm on public datasets.

Learning from Multiple Video Sequences

• Developed a probabilistic framework based on Iterated Conditional Modes to automatically

learn individual signs from video sequences of continuous sign language sentences.

Tools Used: Matlab, C++

WORK EXPERIENCE

Software Engineer: Satyam Computer Services Limited, India

06/2001 - 06/2003

• Worked as a team member in the development of a database backed web application for

handling the entire repair cycle of trailer units. Client: GE-TIP, USA

• Worked as a team member in the development of a web application for monitoring and

updating inspection data of mobile houses collected using handheld devices. Client: GE-

ModSpace, USA

Tools Used: Java Servlets, Oracle, PL/SQL

Undergraduate Intern: Software Technology Parks of India, Hyderabad, India

05/2002 - 07/2002

• Performed a case study of the DATACOM setup at the STPI, Hyderabad.

TEACHING EXPERIENCE

University of South Florida

Instructor:

• Analysis of Algorithms (Undergraduate), Summer 2004

Teaching Assistant:

• Data Structures (Undergraduate), Fall 2006

• Data Structures (Undergraduate), Spring 2004

• Data Structures (Undergraduate), Fall 2003

PUBLICATIONS

• S. Nayak, S. Stagg, P.W. Lau, B. Carragher and C.S. Potter, “Affinity propagation-based

classification applied to single particle analysis and reconstruction”, To be submitted to

Journal of Structural Biology, Feb. 2008.

• S. Nayak, S. Sarkar, and B. Loeding, “Finding Recurrent Patterns from Continuous Sign

Language Sentences for Automated Extraction of Signs”, To be submitted to Computer Vision

and Image Understanding, Feb. 2008.

• S. Nayak, S. Sarkar, and B. Loeding, “Distribution-based dimensionality reduction applied to

articulated motion recognition”, Accepted for publication in Transactions on Pattern Analysis

and Machine Intelligence, Jan. 2008.

• S. Nayak, S. Sarkar, and B. Loeding, A. Karshmer, “Continuous Sign Language

Recognition”, Young Researchers’ Consortium at International Conference on Computers

Helping People with Special Needs, Jul. 2006.

• S. Nayak, S. Sarkar, and B. Loeding, “Unsupervised Modeling of Signs Embedded in

Continuous Sentences”, IEEE Workshop on Vision for Human-Computer Interaction in

conjunction with CVPR, Jun. 2005.

• S. Nayak, S. Sarkar, and K. Sengupta, “Modeling Signs using Functional Data Analysis”,

Indian Conference on Computer Vision, Graphics and Image Processing, Dec. 2004.

RELEVANT GRADUATE COURSE WORKS

Geometrical and Statistical Pattern Recognition, Computer Vision, Digital Image Processing,

Introduction to Theory of Algorithms, Data Mining, Randomized Algorithms, Advanced Linear

Algebra, Operating Systems, Computer Architecture.

RELEVANT COURSE PROJECTS

• Hand tracking: Designed a multi-modal hand tracking algorithm using both pixel intensity

and 3D depth information. A smart scheme was devised to integrate intensity and depth

information to increase the robustness of the system.

• Circle detection: Implemented a Hough Transform based circle detector for microarray

images.

• Optical character recognition: Implemented and tested several classifiers for a scanned

characters dataset, including minimum distance classifier, Bayes moment classifier, k-

Nearest Neighbor classifier, Reduced Nearest Neighbor, and Condensed Nearest Neighbor

classifiers.

TALKS

• Young Researchers’ Consortium at International Conference on Computers Helping People

with Special Needs, July 2006.

• IEEE Workshop on Vision for Human-Computer Interaction in conjunction with CVPR, June

2005.

HONORS

• Selected for the 1st Google Workshop for Women Engineers held at Mountain View, CA, in

January 2006.

• Meritorious Girls Scholarship: Awarded by the State Govt. for earning the 5th position among

all women candidates who appeared for Joint Entrance Examination (1997) for admission

into Engineering Colleges in the state of Orissa, India, 1997-2001.

REFERENCES

Dr. Sudeep Sarkar: Professor, Department of Computer Science and Engineering, University of

South Florida, Tampa, FL. ( sarkar@csee.usf.edu )

Dr. Bridget Carragher: Associate Professor, Automated Molecular Imaging Group, The Scripps

Research Institute, La Jolla, CA. ( bcarr@scripps.edu )

Dr. Barbara Loeding: Associate Professor, Department of Special Education, University of South

Florida, Lakeland, FL. ( bloeding@gmail.com )