光电子学
光电流
宽带
材料科学
光电探测器
异质结
铁电性
计算机科学
延迟(音频)
亮度
量子效率
功率消耗
光学
非线性系统
信号(编程语言)
低延迟(资本市场)
信号处理
电子工程
带宽(计算)
电效率
光通信
还原(数学)
图像传感器
量子阱
功率(物理)
作者
Yuting Ye,Jianmin Wu,Yin Zhang⋆,H F,Qingyan Deng,Jialing Jian,Renjie Tang,Boshu Sun,Yuxiang Zeng,Yinyin Song,Juan Wang,Hongtao Lin,Shiyu Zhao,Lan Li
出处
期刊:Nano Letters
[American Chemical Society]
日期:2026-04-21
卷期号:26 (17): 5644-5651
标识
DOI:10.1021/acs.nanolett.5c06055
摘要
In-sensor computing has emerged as a promising paradigm to overcome power consumption and latency bottlenecks in vision systems. Here, we demonstrate a reconfigurable in-sensor image enhancement strategy based on an In2Se3/PdSe2 ferroelectric heterojunction. The photodetector exhibits a broadband spectral response (400–1550 nm) and a high external quantum efficiency exceeding 104%. By synergistically leveraging electrostatic and ferroelectric fields to tune the band alignment, we achieve programmable carrier collection efficiency, leading to a gate-tunable nonlinear photocurrent response. This hardware-level nonlinearity enables dual imaging modes for adaptive imaging: a low-light signal amplification mode to boost brightness and an overexposure recovery mode to compress contrast. By implementing a programmable photoresponse into a single photodetector, our approach bypasses redundant data transmission, providing a compact and energy-efficient solution for intelligent vision systems.
科研通智能强力驱动
Strongly Powered by AbleSci AI