The laboratory of optics investigates many interesting topics related to optical imaging. These include novel optical and computational techniques for optical diffraction tomography (ODT) through thick samples, imaging through multi-mode fibers, and the printing of optical waveguides.
Optics and learning-based algorithms have a two-way connection such that optical problems were shown to be solved with neural networks efficiently and neural network structures were implemented with optical systems. In our laboratory, we have contributed to both applications, tackling optical problems such as optical diffraction tomography and image transmission through optical fibers with deep neural networks, in addition to realizing neuromorphic computation with multimode optical fibers.
Membrane-less electrolyzers utilize fluidic flow to separate hydrogen and oxygen bubbles. They have a simple design, low ionic resistance, and low maintenance and manufacturing costs. Image from Energy Environ. Sci.,2015,8, 2003.