Adaptive Optoelectronic Transistor Enabled by Nanoporous Patterning for Low‐Contrast Image Recognition

Abstract Photoadaptive neuromorphic device is fundamental for artificial vision system to effectively perceive and accurately process information in complex light scenes. However, adaptive optoelectronic device remains a challenge due to the difficulty in achieving the light‐dependent modulation of charge dynamics. Here, an adaptive optoelectronic transistor enabled by the conformal nanoporous patterning that can perform the biomimetic eye functions and facilitate the precise contrast enhancement under harsh lighting conditions is reported. The enhanced light absorption, photoexciton separation, and charge injection properties are originated from the localized electric field in the nanoporous optoelectronic transistor, which further endow the light intensity‐dependent charge trapping and de‐trapping dynamics, realizing the adaptive decay rate of conduction states. Based on the improved device performances and the adaptive properties, an artificial vision system is constructed to present the biomimetic eye model with visual selective attention functions and a contrast enhancement in the low‐contrast sub‐images. The eye rotation and closing functions, as well as a high recognition accuracy of 95.4% for non‐ideal image with a contrast as low as 0.8% are demonstrated. This study provide an approach for the photoadaptive neuromorphic device and support their visual adaptation applications in biomimetic eye system.