Colloquium: Amit Ashok

    Thursday, September 17, 2015 - 3:30pm - 5:00pm
    Meinel 307

    “Information Optimal Imaging: Detection, Classification and Adaptation


    Natural scenes/objects exhibit smoothness/correlation at multiple scales and across multiple dimensions that leads to compressible representations. This “compressibility” aspect has been very successfully exploited in image/video compression (e.g. JPEG and JPEG2000, MPEG) and more recently in compressive imaging. For specific tasks, such as detection/classification, arguably there exist even more sparse/compressible representations that can and should be exploited for optimal imaging. Information theory provides rigorous and task-based measures to evaluate and optimize measurements that can be used for optimal imaging system design. There are three key challenges associated with information optimal imaging, all of which derive from data dimensionality: (1) object/scene statistical models, (2) scalable information-theoretic metrics, and (3) constrained system optimization. In this talk, I will discuss our recent progress, on all three fronts, in application of information-theoretic metrics to imaging system design for detection and classification tasks, ranging from EO/IR to X-ray. Another related area of work that I will talk about is adaptive imaging. Adaptive measurement design exploits information embedded in past measurements to “adapt” future measurements, which has the potential for significant performance improvements.

    Speaker Bio(s): 

    Amit Ashok an Assistant Professor in the College of Optical Sciences and the Department of Electrical and Computer Engineering (ECE) at the University of Arizona and leads the Intelligent Imaging and Sensing Lab (I2SL). He received his PhD and M.S. degrees in ECE from University of Arizona and University of Cape Town in 2008 and 2001 respectively. His research experience spans both industry and academia and his research interests include computational imaging and sensing, physical optics, Bayesian inference, statistical learning theory and information theory. He has made key contributions in task-based joint-design framework for computational imaging and information-theoretic system performance measures such as the task-specific information. With his multi-disciplinary contributions he has been invited to speak at various OSA, IEEE, SIAM and SPIE conferences.