神经形态工程学
计算机科学
计算机硬件
像素
数字光处理
显示设备
晶体管
人工智能
人工神经网络
电气工程
工程类
投影机
操作系统
电压
作者
Qizhen Chen,Xianghong Zhang,Yaqian Liu,Yujie Yan,Rengjian Yu,Xiumei Wang,Zenan Lin,Huaan Zeng,Lujian Liu,Huipeng Chen,Tailiang Guo
出处
期刊:Nano Energy
[Elsevier]
日期:2022-04-01
卷期号:94: 106931-106931
被引量:19
标识
DOI:10.1016/j.nanoen.2022.106931
摘要
Ultra-high integration, ultra-high resolution, intelligence and human-machine interaction is the inevitable trend of flat panel displays. However, most intelligent display systems are based on traditional Von Neumann architecture, which need to call data frequently because the storage and computation are separated, resulting in serious loss of power consumption and speed. In this paper, the concept of "neuromorphic display" device is proposed for the first time. The display technology is organically combined with the artificial neural network system to build a neuromorphic display network including the sensor module, processing module and display module, which endows the tiny artificial neural network system with the ability of perception, dispose and execution. The neuromorphic display enable to avoid serious loss of power consumption and speed without calling data frequently, and the color of the emitted light directly represents the result of neuromorphic display’s perception and processing of various external environment haptic stimuli. Furthermore, the neuromorphic element is applied in display that one-transistor one-synaptic transistor (1T1S) pixel circuit is demonstrated to replace two-transistor one-capacitance (2T1C) pixel circuit in traditional display to simplify pixel circuit. Therefore, the neuromorphic display showed great potentials in interactive input/control devices, intelligent wallpaper, robots and medical/health monitoring devices, and the 1T1S pixel circuit can simplify circuit design for active matrix display.
科研通智能强力驱动
Strongly Powered by AbleSci AI