Vijayakumar Bhagavatula

U.A. and Helen Whitaker Professor – ECE ; Affiliated Faculty – DSSC ; Associate Dean of Faculty and Graduate Affairs – CIT

Department	Data Storage Systems Center Electrical and Computer Engineering
Office	B205 Hamerschlag Hall
Telephone	(412)-268-3026
Fax	(412)-268-6345
Email	kumar@ece.cmu.edu
Website	http://www.ece.cmu.edu/~kumar/
Assistant	Kara Knickerbocker

Research Interests

Pattern Recognition for Biomerics

One way of matching live biometric signatures (e.g., face images, iris images) with stored templates is to correlate the two. Professor Kumar and his students are developing several spatial frequency-domain methods to perform these correlations in the presence of significant appearance variability (e.g., due to illumination changes, expression differences, etc.) in the biometric signatures. These correlation filters have performed very well in the recent Face Recognition Grand Challenge (FRGC) and Iris Challenge Evaluation (ICE) conducted by US National Institute of Standards and Technology (NIST).

Coding & Signal Processing for Data Storage

As we attempt to increase the densities of data storage systems, bits and tracks come much closer, causing increased intersymbol interference and noise. Advanced coding and signal-processing methods are being developed to provide reliable (i.e., low bit error rate) detection of the digital data stored on various media. Professor Kumar's research considers timing recovery methods and the equalization and detection of data recorded on high-density magnetic tape recording systems, holographic data storage systems and optical storage systems. His group is also investigating the use of low density parity check (LDPC) codes for data storage systems.