Colloquiuim: Michael Gehm

    Date: 
    Thursday, April 28, 2016 - 3:30pm - 5:00pm
    Location: 
    Meinel 307
    Description: 

    Adaptive architectures for spectral processing

    Abstract(s): 

    Introducing adaptive components into the architectures of optical sensor systems facilitates powerful new modes of operation that blend acquisition and exploitation. In this talk I will discuss three spectral applications where my group has experimentally validated gains that arise from such systems—dynamic range matching, direct classification spectrometry, and direct classification spectral imaging—and will discuss the new application we are currently exploring. I will close with a description of a key challenge to further development of these highly-computational adaptive approaches.

    Speaker Bio(s): 

    Dr. Michael E. Gehm is currently an Associate Professor of Electrical and Computer Engineering Department at Duke University, where he leads the Lab for Engineering Non-traditional Sensors (LENS). Dr. Gehm received his Ph.D. in 2003 in Physics from Duke University for fundamental experimental studies of degenerate Fermi gases. From there his interests turned to Computational Sensing and, after a postdoc in the field, he joined the faculty of the University of Arizona, rising to Associate Professor before his move to Duke in 2013. His primary research interest is computational sensing, with specific applications in optics, x-rays, and mass-spectrometry. Additional interests include fundamental optical physics, and the development of 3D printing technologies to directly fabricate functional electromagnetic components.