Umesh Rajashekar's Research

Umesh's Research

" Results! Why, man, I have gotten a lot of results. I know several thousand things that won't work ."

Adaptive Signal Representations for Image Quality Assessment
People: Prof. Zhou Wang & Prof. Eero Simoncelli

We propose a methodology for quantifying spatio-chromatic distortions based on the observation that human judgments of image quality are relatively insensitive to small changes in the viewing or imaging conditions. Stated geometrically, we exploit the fact that distortions along some directions in the vector space corresponding to the full set of color image pixel values are less perceptible than others, and that these directions generally depend on the content of the original image.

Relevant Publications:

U. Rajashekar, Z. Wang, and E. P. Simoncelli,"Perceptual quality assessment of color images using adaptive signal representation", Proc. SPIE, vol.7527(1), pp. 75271L--9,San Jose, California, USA, Jan 2010.

GAFFE: A Gaze-Attentive Fixation Finding Engine
People: Dr. Ian van der Linde, Prof. Al Bovik & Prof. Larry Cormack

Using the DOVES eye movement database, we studied four low-level local image features: luminance, contrast, and bandpass outputs of both luminance and contrast at human fixations. We discovered that image patches around human fixations had, on average, higher values of each of these features than image patches selected at random. Using these measurements, we developed a new algorithm called GAFFE that selects image regions as likely candidates for fixation. These regions are shown to correlate well with fixations recorded from human observers.

Relevant Publications:

U. Rajashekar, I. van der Linde, A. C. Bovik, and L. K. Cormack. "GAFFE: A gaze-attentive fixation finding engine", IEEE Transactions on Image Processing, 17(4):564–573, April 2008.

DOVES: A Database of Eye MovementS
People: Dr. Ian van der Linde, Prof. Al Bovik & Prof. Larry Cormack

DOVES (a Database Of Visual Eye movementS) is a collection of eye movements from 29 human observers as they viewed 101 natural calibrated images. Recorded using a high-precision dual-Purkinje eye tracker, the database consists of around 30,000 fixation points, and is believed to be the first large-scale database of eye movements to be made available to the vision research community.

Relevant Publications:

I. van der Linde, U. Rajashekar, A. C. Bovik, L. K. Cormack. "DOVES: A database of visual eye movements", Spatial Vision, 22(2): 161-177, 2009

Classification Images at Eye Fixations
People: Prof. Al Bovik & Prof. Larry Cormack

The goal of this project was to investigate whether observers used structural cues to deploy their fixations as they searched for simple geometric targets embedded at very low signal-to-noise ratios in 1/f noise. By computing classification images at observers' fixations, we were able to reveal idiosyncratic, target-dependent features used by observers in this visual search task.

Relevant Publications:

U. Rajashekar, A. C. Bovik, and L. K. Cormack, "Visual Search in Noise: Revealing the Influence of Structural Cues by Gaze-contingent Classification Image Analysis", Journal of Vision, Special Issue on Finding visual features: Using stochastic stimuli, Vol: 6 No: 4, March 2006. Page(s): 379-386

SIVA: The Signal, Image, & Video AudioVisualizations
People: George Panayi, Frank Baumgartner, & Prof. Al Bovik

SIVA - The Signal, Image and Video Audiovisual Demonstration Gallery is a collection of didactic tools that facilitate a gentle introduction to concepts in signal and image processing. Equipped with informative visualizations and a user-friendly interface, SIVA uses novel LabVIEW and MATLAB based demonstrations to illustrate the power and beauty of signal and image processing algorithms. The SIVA demonstration gallery is being used at 600+ academic, industrial and other sites around the world.

Relevant Publications:

U. Rajashekar, G. C. Panayi, F. P. Baumgartner, and A. C. Bovik, "The SIVA Demonstration Gallery for Signal, Image, and Video Processing Education", IEEE Transactions On Education, Vol: 45 No: 4, November 2002, Page(s): 323-335

Real-time Foveation: An eye tracker-driven imaging system
People: Prof. Zhou Wang & Prof. Al Bovik

In the Fall of 2001, we built a real-time foveation filtering system to demonstrate the foveation feature of the HVS. The system records observers' gaze direction and automatically 'foveates' the image or video being observed in real time. Since the image/video at the point of gaze is always high-resolution, the foveated display is perceptually indistinguishable from the original to the observer!