Novel Optoelectronic Devices

 In our hyper-connected society, vast amounts of unstructured image data are being generated from various sources such as augmented-reality (AR) displays, IoT, robots, autonomous vehicles, etc. The huge amount of image data needs to be processed efficiently and quickly to enhance user experience, however, traditional von-Neumann architecture for image data storage and processing has fundamental limitations in power efficiency, data processing speed, and high-density integration. Therefore, there is an urgent need for developing new technologies to overcome this challenge.

 

 In our lab, we are focusing on the development of a new optical neuromorphic and in-sensor computing devices, which are capable of simultaneously performing computation, storage, and on-chip learning within the device itself with both light and electrical signals, enabling high-accuracy image inference. In our new devices, colloidal semiconductor quantum dots and perovskite nanocrystals, which have excellent light absorption and charge transport properties, are used as a light-sensing layer. We believe our in-sensor computing devices will provide a solution for the image data processing problems, offering a new paradigm in semiconductor industry as well as other fields of study including artificial intelligence, optical quantum computing, hyperspectral imaging, and optoelectronic integrated circuits.