A color enhancement maps colors to enhanced colors. Simple enhancements, like brightening all colors, are easy to implement with today's tools. Our work considers how to enable designers to create custom color enhancements by learning a look-up-table that specifies how colors should be transformed. A look-up table can be estimated from a few example color pairs given by a designer or from enhanced images or videos. The estimated look-up table can be stored as an ICC profile, and then images (or video) can be enhanced using the standard color management modules that are already built-into hardware and software systems. Our work focuses on the estimation problems that arise. In particular, we investigate new neighborhood definitions for local learning that can lead to more robust estimation and allow small neighborhoods for flexibility. Our papers cover both custom color enhancements and gamut expansion.

Personnel:

Maya R. Gupta (EE Associate Professor)

Eric Garcia (EE Ph.D Student)

Jayson Bowen (now: UW EE Masters Student)

Andrey Stroilov (now at Google)

Hyrum Anderson (EE PhD Student)

Publications:

"Learning custom color transformations with adaptive neighborhoods," Maya R. Gupta, Eric K. Garcia, and Andrey Stroilov, Journal of Electronic Imaging (with Open Access), vol. 17, no. 3, 2008. (Experimental Data)

"Gamut Expansion for Video and Image Sets," Hyrum Anderson, Eric K. Garcia, and Maya R. Gupta, Computational Color Imaging Workshop, 2007.

"Custom Color Enhancements by Statistical Learning," Maya R. Gupta, Proceedings of the IEEE Intl. Conf. on Image Processing, pp. 968-971, 2005.

"Simulating the effect of illumination using color transformations," Maya R. Gupta, Steve Upton, and Jayson Bowen, Proceedings of the SPIE Conference on Computational Imaging III, vol. 5674, pp. 248-258, 2005.

Related Work:

Color Processing for Color Reproduction