Optical Neural Networks: Principles, Challenges, and Future Prospects in Computing and Astrophotonics

The rapid development of optical neural networks (ONNs) has introduced new research avenues for enhancing computing power. Owing to the characteristics of optical signals, such as low power consumption, low latency, high parallelism, and large bandwidth, optical computing based on neural network architectures shows promise for processing spatial signals, temporal signals, and on-chip information. However, ONN-based computing lacks a unified architecture. Because of its physical limitations, ONN computing entails different requirements for structure, devices, integration, and signal processing methods across different applications. This review addresses the practicality of ONNs, introduces the fundamental principles of ONN-based computing, describes the challenges in different application scenarios, and highlights the crucial role of optical computing in our future information society.