HRISHIKESH KARVIR

*** ******** *******, ********, ** 19446

TEL: 937-***-****, E-MAIL: abmqfn@r.postjobfree.com

CAREER OBJECTIVE

Seeking a research scientist position with an organization that fosters innovation and provides leadership

opportunities. My background in machine learning, along with work experience at different project life-cycle

stages, will be a complete and valuable asset for the organization’s current goals and future directions.

EDUCATION

Wright State University, Dayton, OH

Ph.D. in Engineering [Sensor Signal and Image Processing] 03/2006 – Present

M.S. in Engineering [Biomedical Engineering] 08/2003 –

05/2006

Mumbai University, Mumbai, India

B.E. in Biomedical Engineering 06/1998 – 06/2002

RESEARCH EXPERIENCE

Ph.D. Dissertation: Design, Development and Validation of a Data- and 03/2006 – Present

Feature-level Information Fusion System

Sponsor: The Leonard Wood Institute, Fort Leonard Wood, MO

My dissertation research aims at evaluation of information fusion and machine learning algorithms. A novel

Convex Hull-based pattern classifier scheme was developed and compared to traditional classifiers.

Developed a software framework for optimized feature discovery, information fusion and evaluation of

different machine learning algorithms used in computer-aided detection applications

Applied the framework for image data acquisition, noise filtering, image registration, segmentation,

feature discovery, data manipulation and feature extraction/selection for pattern classification

Evaluated techniques such as Principle Component Analysis (PCA) and Linear Discriminant Analysis

(LDA) for feature selection from a diverse and extensive feature set

Conceptualized a novel, morphing Convex Hull-based pattern classifier approach for improving

prediction accuracy in multivariate and mixed probability density distribution environments

Optimized the Convex Hull classifier parameters using Monte Carlo simulations and nonlinear

regression modeling

Analyzed performance of the Convex-Hull approach against quadratic Bayes, k-nearest neighbors (k-

NN) and support vector machine (SVM) algorithms using cross-validation techniques

Investigated parametric and nonparametric multivariate density estimation techniques for score-based

evaluation and confidence assessment of detections

Multi-modal Imaging System for Ground-Based Target Detection 11/2007 – 12/2008

Sponsor: The Leonard Wood Institute, Fort Leonard Wood, MO

1

This research project involved developing a portable and scalable multi-sensor/multi-spectral test bed for

target detection and imaging applications.

Designed and integrated a portable and expandable hardware platform for interfacing multi-sensor/multi-

spectral imagers, using commercial, off-the-shelf components

Deployed a vehicle mounted multi-spectral prototype unit that uses an electro-optical camera and a short-

wave infrared camera for visual cueing of improvised explosive devises (IEDs)

Developed a library of routines for extracting statistical, structural and intensity-based features targeted

towards identifying objects of interest

Developed a library of routines for lens distortion correction, noise removal, mutual information-based

image registration, nuisance detection and nuisance suppression for IED target detection

Nonlinear Enhancement for Visual Target Detection 09/2005 – 11/2006

Sponsor: Applied Imaging Sciences, Dayton, OH, (subcontracted from Air Force Office of Sponsored

Research)

Designed an enhanced decision-level target detection scheme based on nonlinear weighted

voting of multiple target detection algorithms

Acquired test and training images to simulate aerial imagery using reduced scale terrain boards

and target replicas of scud missile launchers, military aircraft and transport vehicles

Developed indigenous algorithms that exploit unique power spectral signatures and spatial

intensity distributions of targets, thereby providing high detection sensitivity and specificity

Determined optimal algorithm parameters and assessed performance of the pattern recognition

schemes through receiver operating characteristics (ROC) curve analyses

Master’s Thesis: Digital Radiographic Absorptiometry: Validation 01/2004 – 01/2006

Through Detective Quantum Efficiency Analysis

Sponsor: Wright State University, Dayton, Ohio

This research work was aimed at quantitative performance assessment of x-ray films and digital sensors for

bone mineral density (BMD) evaluation, under varying clinical operating conditions, using standard dental x-

ray equipment.

Performed dental x-ray tube characterization, using a NaI-based spectrometer and radiation dose meter,

for accurate x-ray film and digital sensor evaluation

Quantitatively compared performances of a CMOS sensor and a dental x-ray film through modulation

transfer function (MTF), noise power spectrum (NPS) and detective quantum efficiency (DQE) analysis

Studied the impact of changing exposure on MTF, and subsequently on sensor and film DQE values.

TEACHING EXPERIENCE

Guided undergraduate students in their summer project “Multimodal anomalous 06/2009 – 08/2009

vehicle detection.” sponsored by TechEdge, Dayton, Ohio

Supervised a graduate student project, “Design of a mobile cart and a mechanical 04/2008 – 11/2008

camera mounting chassis for a portable imaging system.”

Graduate teaching assistant in the Department of Biomedical, Industrial and 08/2004 – 07/2008

Human Factors Engineering for the following courses:

2

Photon Radiation (BME 470/670), Medical Imaging (BME 471/671), Instrumentation for Radiation

Measurement (BME 737) and Medical Image Processing (BME 734)

PUBLICATIONS

Karvir HV, Skipper JA, Repperger DW, “Entropy selective mutual information-based image

registration,” National Aerospace Electronics Conference, Fairborn, OH, July 2008.

Karvir HV and Skipper JA, “Power spectrum weighted edge analysis for target detection in images,”

Proceedings of the 2007 SPIE Defense and Security Symposium, Orlando, FL, April 2007.

Karvir HV and Skipper JA, “Multi-sensor Integration Platform for System Evaluation and Fusion

Algorithm Assessment,” 5th Annual Dayton Engineering and Sciences Symposium, Dayton, OH, October

2009. (Abstract)

Karvir HV, Skipper JA and Repperger DW, “Power spectrum weighted edge analysis for target

detection in images,” 2nd Annual Dayton Engineering and Sciences Symposium, Dayton, OH, October

2006. (Abstract)

Karvir HV and Skipper JA, “Detective quantum efficiency comparison of radiographic film versus a

CMOS sensor for radiographic absorptiometry based BMD assessment,” Dentomaxillofacial Radiology

(Submitted).

PRESENTATIONS

Karvir HV and Skipper JA, “Multi-sensor Integration Platform for System Evaluation and Fusion

Algorithm Assessment,” 5th Annual Dayton Engineering and Sciences Symposium, Dayton, OH, October

2009.

Karvir HV, Skipper JA and Repperger DW, “Power spectrum weighted edge analysis for target

detection in images,” 2nd Annual Dayton Engineering and Sciences Symposium, Dayton, OH, October

2006.

Karvir HV and Skipper JA, “Digital radiographic absorptiometry: validation through detective

quantum efficiency analysis,” IEEE Symposium (Dayton Chapter), Dayton, OH, April 2007.

Semi-annual technical presentations on research work at the BioMedical Imaging Laboratory

seminars

MEMBERSHIPS AND AWARDS

Tau Beta Pi Engineering Honor Society 01/2005 –

Present

Alpha Eta Mu Beta Biomedical Engineering Honor Society 03/2008 –

Present

Society of Photographic Instrumentation Engineers (SPIE) 03/2007 –

Present

Toastmasters International 04/2009 –

Present

Volunteer, Miami Valley Hospital, Dayton, Ohio 01/2004 –

06/2007

Ph.D. in Engineering Assistantship 06/2009 –

05/2010

3

College of Engineering and Computer Science Scholarship 08/2005 –

07/2007

RELEVANT COURSE WORK

Core Courses Core Courses Breadth Courses Miscellaneous

Data Mining Photon Radiation NMR Spectroscopy Medical Image Processing

Digital Image Processing Medical Imaging Medical Ultrasonics Systems Engineering

Computer Vision I & II Computed Tomography Instrumentation for Design and Analysis of

Radiation Measurement Experiments

Pattern Classification Digital Signal Processing

COMPUTER SKILLS

Programming Languages: Matlab, LabVIEW, C, Weka, R

Statistical Packages: JMP

Image Processing Packages: Image Pro-Plus, Adobe Photoshop

Applications: Microsoft Word, Excel, Front page

Operating Systems: Windows XP, Vista, Windows Server 2003

4

Contact this candidate