Dejian Zhang | Searcharxiv

arXiv2024

Optoelectronic generative adversarial networks

Artificial intelligence generative content technology has experienced remarkable breakthroughs in recent years and is quietly leading a profound transformation. Diffractive optical networks provide a promising solution for implementing generative model with high-speed and low-power consumption. In this work, we present the implementation of a generative model on the optoelectronic computing architecture, based on generative adversarial network, which is called optoelectronic generative adversarial network. The network strategically distributes the generator and discriminator across the optical and electronic components, which are seamlessly integrated to leverage the unique strengths of each computing paradigm and take advantage of transfer learning. The network can efficiently and high-speed process the complex tasks involved in the training and inference of the generative model. The superior performance of these networks is verified by engaging three types of generative tasks, image generation, conditional generation, and image restoration. By synergistically combining the strengths of optical and electronic computing, the optoelectronic generative adversarial network paves the way for the development of more powerful and accessible artificial intelligence generative content technology that can unlock new creative possibilities across a wide range of applications.

physics.optics↗

arXiv2023

Decision-making and control with diffractive optical networks

The ultimate goal of artificial intelligence is to mimic the human brain to perform decision-making and control directly from high-dimensional sensory input. Diffractive optical networks provide a promising solution for implementing artificial intelligence with high-speed and low-power consumption. Most of the reported diffractive optical networks focus on single or multiple tasks that do not involve environmental interaction, such as object recognition and image classification. In contrast, the networks capable of performing decision-making and control have not yet been developed to our knowledge. Here, we propose using deep reinforcement learning to implement diffractive optical networks that imitate human-level decision-making and control capability. Such networks taking advantage of a residual architecture, allow for finding optimal control policies through interaction with the environment and can be readily implemented with existing optical devices. The superior performance of these networks is verified by engaging three types of classic games, Tic-Tac-Toe, Super Mario Bros., and Car Racing. Finally, we present an experimental demonstration of playing Tic-Tac-Toe by leveraging diffractive optical networks based on a spatial light modulator. Our work represents a solid step forward in advancing diffractive optical networks, which promises a fundamental shift from the target-driven control of a pre-designed state for simple recognition or classification tasks to the high-level sensory capability of artificial intelligence. It may find exciting applications in autonomous driving, intelligent robots, and intelligent manufacturing.

cs.LG↗