作者
Kexin Miao,Qinhao Lin,Tao Zhao,X. Xu
摘要
ABSTRACT Photodetectors are evolving toward next‐generation systems with intrinsic intelligence, where sensing is tightly coupled with information encoding, real‐time processing, and adaptive control at the device or sensor‐system level. In this context, intelligence refers to architecture‐level capabilities that enable dynamic regulation of operational parameters, efficient optical information processing, and robust performance under diverse environments. By integrating tunable materials, reconfigurable architectures, and embedded or near‐sensor computing, advanced smart photodetectors achieve enhanced sensitivity, multifunctionality, reduced data redundancy, and low‐latency responses. The synergy between photodetection and artificial intelligence further enables in‐sensor tasks such as image recognition, target tracking, and pattern classification, supporting rapid decision‐making at the sensing front end. These capabilities are critical for emerging applications in medical diagnostics, autonomous systems, defense, and quantum communication. Despite significant progress, challenges remain in material integration, fabrication scalability, stability during functional reconfiguration, high‐speed multi‐channel processing, and long‐term reliability. This review classifies advanced smart photodetectors into five representative categories, each representing a distinct pathway toward sensor‐level intelligence, and analyzes their mechanisms, information‐processing strategies, performance trade‐offs, and application‐oriented optimizations, providing guidance for future intelligent photodetection systems.