CCM Colloquium: David Barmherzig (Toronto Computational Imaging Group)

Wednesday Oct 30, 2024, 10:00 AM → 11:00 AM America/New_York

3rd floor classroom (162)

Presenter: David Barmherzig ( Toronto Computational Imaging Group)

Title: Ultra-fast and ultra-wideband imaging via photon detection and statistical estimation

Abstract: Recent technological advances have enabled imaging methods that can capture ultra-fast events, i.e. events occurring at light-speed timescales, including the propagation of light itself. Indeed, the 2023 Nobel Prize in Physics was awarded for "attosecond physics", an emerging new field of multidisciplinary study that is only possible via ultra-fast imaging. Current ultra-fast imaging methods are enabled by sophisticated, hardware-driven setups which can only record at pre-determined speeds which must synchronized with external light pulses. As such, they cannot identify ultra-fast phenomena occurring spontaneously within a larger event duration. 
In this talk, a new method for ultra-fast imaging is introduced, which is both: (i) computationally-driven, and does not require external, synchronized light sources, and (ii) "ultra-wideband", i.e. it can simultaneously image across multiple timescales and "zoom in" to detect ultra-fast phenomena. This new imaging method is based on reconstructing images from a stream of individual photon arrival times via statistical estimation and/or inverse problem techniques.

Bio: David Barmherzig is a research scientist with the Toronto Computational Imaging Group, based at the University of Toronto. He previously completed his Ph.D. at Stanford University and was a Research Fellow at the Flatiron Institute Center for Computational Mathematics (CCM). His research interests include computational mathematics and imaging, signal processing, deep learning, statistics and physics.