Scientific Lectures //
DRASTIC: Dynamically Recongurable Architecture Systems for Time-varying Image Constraints
Marios S. Pattichis-Department of Electrical and Computer Engineering at the University of New Mexico
Presented: September 16, 2014
ABSTRACT: The overwhelming majority of image processing algorithms and architectures are designed to satisfy fixed performance constraints. Yet, time-varying constraints can be introduced by available energy (e.g., mobile devices), available wireless network bandwidth, user constraints, or signal content itself. This talk will introduce a multi-objective optimization framework for modeling jointly-optimal architectures and algorithms subject to time-varying constraints. The solutions to such problems can be realized through the development of Dynamically Reconfigurable Architecture Systems for Time-varying Image Constraints (DRASTIC). Example of DRASTIC systems and applications in mobile video communications, image and video analysis will be discussed in this talk.