记忆电阻器
神经形态工程学
冯·诺依曼建筑
计算机科学
感知
钥匙(锁)
适应(眼睛)
建筑
计算机体系结构
人机交互
领域(数学)
系统工程
人工智能
组分(热力学)
纳米技术
开放式研究
电阻随机存取存储器
工程类
智能决策支持系统
作者
Jinlong He,Bo Yang,Bai Tao Sun,Qi Chen,Chuncai Kong,Zelin Cao,Kaikai Gao,Mengna Wang,Longhui Fu,Kun Wang,Teng Wu,Song Ling Wang,Ye Tang,Jinyou Shao
摘要
ABSTRACT With the continuous advancement of artificial intelligence, human‐computer interaction (HCI) is evolving toward intelligent perception, collaborative decision‐making, and natural integration. However, the physical separation of memory and computing units in traditional von Neumann architecture poses critical bottlenecks in terms of power consumption and latency, which limits its ability to meet the requirements of next‐generation HCI systems. Memristors, with their unique advantages of in‐memory computing, synaptic plasticity emulation, and multi‐physical responsiveness, have emerged as a key enabling technology for brain‐inspired perception and interaction systems. This article systematically reviews the latest research progress of memristors in intelligent perception and neuromorphic systems for next‐generation HCI. First, starting from the material system, performance regulation, and application adaptation are analyzed. Second, the strategy of memristors in device design and integration architecture is discussed. Furthermore, the biomimetic perception system based on memristors is emphasized, demonstrating its integration capability at the hardware level of brain‐inspired perception information processing. Finally, it systematically summarizes the research achievements of memristors in the field of HCI applications and discusses the key challenges currently faced. Looking forward, this review suggests that memristors are expected to provide a solid foundation for building high‐performance, low‐power, and biocompatible next‐generation HCI systems.
科研通智能强力驱动
Strongly Powered by AbleSci AI