RAVI GARG

Graduate student

Dept. of Electrical & Computer Engineering

KEB 2244

University of Maryland

College Park, MD 20742

Phone: +1-301-405-3342

email: ravig AT umd DOT edu

Email Me

Welcome


Hello ! Welcome to my Electronic home ! Browse through this website to find more about my research and other activities.


Currently, I am a 5th year graduate student at Department of Electrical & Computer Engineering, University of Maryland, College Park, USA. I work as a research assistant under Dr. Min Wu, who leads the Multimedia and Security Team (MAST). My research interests are in broad areas of signal processing, Image & video processing, machine learning, computer vision, and wireless communications. I am looking for job opportunities in the area of digital signal processing, image & video processing, computer vision, and machine learning.

      The area of my current research focus is on audio and video forensics using a power network induced signatures, also known as Electrical Network Frequency (ENF) signals. ENF is the supply frequency in power distribution network, varying with time around its nominal value of 60 Hz (in north America), and 50 Hz in other parts of the world. These variations make the ENF signal a randomly varying signal which is unique for a particular time segment across the interconnected power grid. These signatures are induced in multimedia recordings from a wide variety of sources. e.g., electromagnetic induction from nearby power lines, audible “hum” from electrical equipments, and invisible “flickering” of indoor lights. We develop algorithms to extract and estimate these signals, and use them in a wide variety of multimedia forensics applications such as determining the time of creation of a recording, and determining editing operations such as clip insertion and clip deletion. Recently, we are exploring the capabilities of ENF based forensic technology in determining the location of recording.

      During my summer internship at Mitsubishi Electric Research Labs, Cambridge from May 2012-Aug 2012, I worked in the area of secure biometrics under the guidance of Dr. Shantanu Rane. We designed a new scheme for feature extraction from fingerprints applicable to processing in the encrypted domain.

     Previously, I worked on the problem of localization in wireless sensor networks under the presence of active adversaries who modifies the data transmitted to the localizing node. We proposed an iterative and computationally efficient algorithm based on gradient descent approach to solve the problem of localization in static and mobile wireless sensor networks.

       In my graduate coursework, I have worked on a wide variety of projects in computer vision and machine learning. I designed an image search system using a hierarchical approach of combining color histogram and texture features for first level of search, and using Scale Invariant Feature Transform (SIFT) features for refined search. I have also worked on various machine learning techniques such as PCA, LDA, sparse representation, and support vector machines (SVM) for face recognition and gender identification. I also developed a sensor network localization system using support vector machines by casting the problem in a classification and a regression framework.

    Prior to coming to US for my graduate study, I graduated from Indian Institute of Technology Madras (IIT Madras) in 2008 with a Bachelor of Technology (B.Tech) in Electrical Engineering.  For my senior year thesis, I worked on the problem developing algorithm for  robust speaker recognition under real world noise conditions, under the guidance of Dr. Hema A. Murthy. We used a MFCC based sub-band recombination scheme to select the best features for each user, and learn the user model using  Gaussian mixture modeling.

        My detailed resume can be found here.



News:

  1. 1.Attended APSIPA Annual Summit & Conference in LA to present my work on ENF based localization (Dec 2012)

  2. 2.Won Jimmy H. Lin award for the invention (Aug 2012)

  3. 3.Advanced to PhD candidacy (May 2012)

  4. 4.Won 2011 UMD Invention of the Year award (Apr 2012)

  5. 5.Will be interning at Mitsubishi Electric Research Labs, Cambridge in the summer of 2012.

  6. 6.Won 2012-2013 James A Clark school of Engineering Future Faculty Fellowship. (Jan 2012)

  7. 7.Won Best Student Paper Award in ACM Multimedia 2011 held in Scottsdale, AZ for my research on ENF based multimedia forensics. (Nov. 2011)

  8. 8.Work on secure localization in wireless sensnor networks got accepted for publications in the forthcoming issue of the IEEE transactions on Information Forensics & Security. (Nov. 2011)

Research Interests: Multimedia Signal Processing, Multimedia Forensics, Information Security, Image & video processing, Speech processing, Machine Learning, COmputer Vision, Wireless Sensor Networks,  wireless communication, Quantitative Finance, Economics, Operation Research.