Gate-tunable van der Waals heterostructure for reconfigurable neural network vision sensor
Chenyu Wang(Collaborative Innovation Center of Advanced Microstructures), Shi‐Jun Liang(Collaborative Innovation Center of Advanced Microstructures), Shuang Wang(Collaborative Innovation Center of Advanced Microstructures), Pengfei Wang(Collaborative Innovation Center of Advanced Microstructures), Zhu‐An Li(Collaborative Innovation Center of Advanced Microstructures), Zhongrui Wang(University of Massachusetts Amherst), Anyuan Gao(Collaborative Innovation Center of Advanced Microstructures), Chen Pan(Collaborative Innovation Center of Advanced Microstructures), Chuan Liu(Nanjing University), Jian Liu(Nanjing University), Huafeng Yang(Nanjing University), Xiaowei Liu(Collaborative Innovation Center of Advanced Microstructures), Wenhao Song(University of Massachusetts Amherst), Cong Wang(Collaborative Innovation Center of Advanced Microstructures), Bin Cheng(Collaborative Innovation Center of Advanced Microstructures), Xiaomu Wang(Nanjing University), Kunji Chen(Nanjing University), Zhenlin Wang(University of Massachusetts Amherst), Kenji Watanabe(National Institute for Materials Science), Takashi Taniguchi(National Institute for Materials Science), J. Joshua Yang(University of Massachusetts Amherst), Feng Miao(Collaborative Innovation Center of Advanced Microstructures)
Early processing of visual information takes place in the human retina. Mimicking neurobiological structures and functionalities of the retina provides a promising pathway to achieving vision sensor with highly efficient image processing. Here, we demonstrate a prototype vision sensor that operates via the gate-tunable positive and negative photoresponses of the van der Waals (vdW) vertical heterostructures. The sensor emulates not only the neurobiological functionalities of bipolar cells and photoreceptors but also the unique connectivity between bipolar cells and photoreceptors. By tuning gate voltage for each pixel, we achieve reconfigurable vision sensor for simultaneous image sensing and processing. Furthermore, our prototype vision sensor itself can be trained to classify the input images by updating the gate voltages applied individually to each pixel in the sensor. Our work indicates that vdW vertical heterostructures offer a promising platform for the development of neural network vision sensor.