Dehui Zhang is a postdoctoral researcher in Electrical Electrical Engineering and Computer Sciences at the University of California, Berkeley and a research affiliate in the Materials Science Division at Lawrence Berkeley National Laboratory. He received a Ph.D. in Electrical and Computer Engineering from University of Michigan, Ann Arbor in 2021, and was a postdoctoral researcher at University of California, Los Angeles in 2021-2023. Dehui Zhang joined Javey Research Group in September 2023.
Research
Multi/hyperspectral imaging collects spatial and spectral information as 3D hypercubes, which can be post-processed to identify materials and structures. However, the dense hypercube data limits the frame rate, resolution, and power efficiency of spectral machine vision. We explore intelligent sensors that perform machine learning analysis of spectra and compress the result directly as the readouts, thereby lowering the energy consumption and increasing the frame rate by orders of magnitudes. We also explore in-situ training algorithms that enable the sensors to learn from examples. Such intelligent sensors are ideal for inspecting complex, uncalibrated environments with high speed, good portability, and low power budget.
