Vishnu Naresh Boddeti
Contact Electrical and Computer Engineering 412-***-****
Information Carnegie Mellon University abqgo7@r.postjobfree.com
www.ece.cmu.edu/ vboddeti
Pittsburgh, PA 15213
Broadly interested in Computer Vision, Signal Processing and application of Statistical Learning
Research
methods in these disciplines. Speci cally object and activity recognition, biometrics recognition
Interests
and data-mining and modeling of time-series data.
Carnegie Mellon University, Pittsburgh, USA
Education
Ph.D, Electrical and Computer Engineering August 2007-Present
Advisor: Prof. Vijayakumar Bhagavatula
M.S, Machine Learning Department January 2011-Present
Advisor: Prof. Byron Yu
M.S, Electrical and Computer Engineering August 2007-May 2009
Advisor: Prof. Vijayakumar Bhagavatula
Indian Institute of Technology, Madras, India
BTech, Electrical Engineering August 2003-July 2007
Advisor: Prof. A N Rajagopalan
B.V.K. Vijaya Kumar, Vishnu Naresh Boddeti, Jonathan M. Smereka, Jason Thornton and Marios
Book Chapters
Savvides, Application of Bayesian Graphical Models to Iris Recognition, submitted to Handbook
of Statistics, vol. 31 (eds: C.R. Rao and V. Govindaraju), May 2011
Raghavender Jillela, Arun Ross, Vishnu Naresh Boddeti, B. V. K. Vijaya Kumar, Xiaofei Hu,
Robert Plemmons, Paul Pauca, An Evaluation of Iris Segmentation Algorithms in Challenging
Periocular Images, in Handbook of Iris Recognition, K. Bowyer and M. Burge (Eds.), Springer
2012 (to appear)
Andres Rodriguez, Vishnu Naresh Boddeti, B.V.K Vijaya Kumar and Abhijit Mahalanobis, Max-
Journal
imum Margin Correlation Filter: A New Approach for Simultaneous Localization and Classi ca-
Publications
tion, IEEE Transactions on Image Processing, Under Review
Vishnu Naresh Boddeti and B.V.K Vijaya Kumar, A Framework for Binding and Retrieving Class-
Speci c Information to and from Image Patterns using Correlation Filters, IEEE Transactions
on Pattern Analysis and Machine Intelligence, Under Review
Vishnu Naresh Boddeti and B.V.K Vijaya Kumar, Extended Depth of Field Iris Recognition using
Unrestored Wavefront-Coded Imagery, IEEE Transactions on Systems, Man, and Cybernetics -
Part A (SMC-A), May 2010.
Arun Ross, Raghavender Jillela, Jonathon M. Smereka, Vishnu Naresh Boddeti, B. V. K. Vijaya
Conference
Kumar, Ryan Barnard, Xiaofei Hu, Paul Pauca, Robert Plemmons, Matching Highly Non-Ideal
Publications
Ocular Images: An Information Fusion Approach, 5th IAPR International Conference on Bio-
metrics, 2012 (oral).
Vishnu Naresh Boddeti, B.V.K. Vijaya Kumar and Krishnan Ramkumar, Improved Iris Segmen-
tation Based on Local Texture Statistics, 45th Asilomar Conference on Signals, Systems and
Computers, 2011. (oral, invited paper)
Vishnu Naresh Boddeti, Jonathan Smereka and B.V.K. Vijaya Kumar, A comparative evalu-
ation of iris and ocular recognition methods on challenging ocular images, International Joint
Conference on Biometrics, 2011. (oral)
Vishnu Naresh Boddeti, Fei Su and B.V.K. Vijaya Kumar, A Biometric Key-Binding and Tem-
plate Protection Framework using Correlation Filters, Proceedings of the Third International
Conference on Advances in Biometrics (ICB), pp. 919-929, 2009.
Vishnu Naresh Boddeti and B.V.K. Vijaya Kumar, Extended Depth of Field Iris Recognition
with Correlation Filters, Biometrics: Theory, Applications and Systems (BTAS), 2008. (oral,
nominated for best-paper award)
Carnegie Mellon University, Pittsburgh, USA(August 2009 - Present)
Current Projects
Correlation Filtering Theory for large scale biometric recognition and object detection. Max-
imum Margin approach to designing classi ers for translation, scale and rotation invariant
simultaneous object detection and tracking with extensions to action recognition in videos.
Discriminative manifold learning for simultaneous super-resolution and recognition of faces
and textures at very low-resolutions (8x8 pixels).
Probabilistic Deformation Models for Face and Ocular Recognition.
Time-Series data-mining and analysis with applications to management of data-centers (Hadoop
clusters) and modeling population neural spike counts for neural prosthetics.
Biometric Template Protection and Key Binding. The goal is to combine biometric recognition,
biometric template protection and key-binding into one framework using Correlation Filters.
We proposed a robust algorithm for retrieving cryptographic keys (upto 600 bits) with faces
and palmprints with a 1% error on the Multi-PIE face database and PolyU palmprint database.
This technique was also extended to multi-modal and multi-class scenarios. Also working on
extensions to binding and retrieving information to and from actions sequences.
Carnegie Mellon University, Pittsburgh, USA
Past Projects
Extended Depth of Field Iris Recognition with Correlation Filters and Wavefront Coding
achieving a depth of eld improvement by a factor of up to 5 over a conventional camera.
Region based active contour segmentation for iris recognition.
Carnegie Mellon University, Pittsburgh, USA
Course Projects
Multi scale dictionary learning for simultaneous image super-resolution and recognition.
Object Classi cation with Spatial Pyramid Matching and Spectral Analysis. Here we combine
spatial domain features (SIFT) with spectral analysis for scene and object classi cation using
SVMs with a spatial pyramid match kernel.
Unsupervised Object Discovery in large databases using topic discovery models from statistical
text literature. Looked into ways in which segmentation can be used to help discover more
meaningful topics.
Feature selection for cognitive state classi cation using MEG data.
Indian Institute of Technology Madras, Chennai, India(August-May, 2007)
Learning based Super-Resolution of face videos using a 3-D Markov Random Field that encodes
the spatial-temporal consistencies and the image formation and degradation process.
Modeling and analysis of a practical reverberation model. Also investigated the e ects of
common audio post processing e ects when applied to human voice and music.
Designed a power electronics based controller for a prototype experiment to measure the fre-
quency of oceanic waves. The waves were generated arti cially in the laboratory.
Ittiam Systems Pvt. Ltd, Bangalore, India (May-July, 2006)
Work Experience
Investigated and implemented (C++) many image denoising algorithms on both still images and
videos with main emphasis on wavelet based Shrinkage denoising methods.
Statistical Machine Learning Physics Based Methods in Vision
Coursework
Learning Based Methods in Vision Optimization
(AT CMU)
Computer Vision Methods of Optimization
Geometry Based Methods in Vision Machine Learning
Applied Stochastic Processes Probablistic Graphical Models
Intermediate Statistics Wavelets and Multi-resolution Techniques
Pattern Recognition Theory Advanced Digital Signal Processing
Algorithms in the Real World Error Control Coding
Teaching Assistant - Digital Signal Processing. Spring 2009
Teaching
Responsibilities include conducting recitations, writing problem solutions and grading exams.
Experience
Teaching Assistant - Signals and Systems. Fall 2009
Responsibilities include conducting recitations, preparing homeworks, preparing exams, preparing
and conducting labs.
Programming: Python, C/C++, Matlab.
Computer
Libraries: Numerical Python (Numpy, Scipy, CVXOPT etc.), OpenCV, Eigen
Skills
Publishing: L TEX
A
Platforms: Various GNU/Linux Distributions, Mac OS X, Microsoft Windows.
Honours and
Doctoral Consortium Fellowship at BTAS 2011
Awards
Dean s Fellowship, Carnegie Mellon University, 2007-Present
Merit Scholarship, Indian Institute of Technology, Madras, 2003-2007
Pratibha Scholarship for Outstanding Academic Achievement, Government of Andhra Pradesh,
India, 2000-2002
Professional
Student Member: IEEE
Activities
Reviewer:
IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI)
International Conference on Biometrics: Theory, Applications and Systems (BTAS)
Leadership
Member: Org Management Steering Committee at CMU (2011-2012).
President: Indian Graduate Student s Association at CMU (2010-2011).
Treasurer: Indian Graduate Student s Association at CMU (2009-2010).
Available on request.
Referees