Adaptive Vision Sensor Based on Nonlinear Negative Photoconductivity Behavior

Abstract

In-sensor adaptive visual systems represent a promising technology applicable across various fields. This method significantly enhances image quality while reducing system complexity, thereby holding substantial scientific significance and practical applications. This study emulates a light-triggered depolarization neuromorphic response utilizing an In₂O₃/C8-BTBT heterojunction transistor device equipped with ion-gel gating. When the heterojunction device is exposed to UV light, electrons in the In₂O₃ layer recombine with holes in the C8-BTBT layer, leading to a rapid decrease in photocurrent and resulting in a significant negative photoresponse. The device is capable of simulating spike-dependent inhibitory currents and multilevel storage capabilities. Moreover, the proposed device is employed in constructing a UV-adaptive retina, facilitating in-sensor adaptive computational imaging by leveraging its unique dependence on UV intensity and temporal characteristics, thereby significantly enhancing the visualization of image details.