Ultracompact meta-imagers for arbitrary all-optical convolution

• Published in: Light, science & applications Vol. 11; no. 1; p. 62
• Main Authors: Fu, Weiwei, Zhao, Dong, Li, Ziqin, Liu, Songde, Tian, Chao, Huang, Kun
• Format: Journal Article
• Language: English
• Published: London Nature Publishing Group UK 18.03.2022; Springer Nature B.V; Nature Publishing Group
• Subjects: 639/624/399/1015; 639/624/400/1103; Applied and Technical Physics; Artificial intelligence; Atomic; Classical and Continuum Physics; Information processing; Lasers; Molecular; Neural networks; Optical and Plasma Physics; Optical Devices; Optics; Photonics; Photons; Physics; Physics and Astronomy
• ISSN: 2047-7538; 2095-5545; 2047-7538
• DOI: 10.1038/s41377-022-00752-5

Electronic digital convolutions could extract key features of objects for data processing and information identification in artificial intelligence, but they are time-cost and energy consumption due to the low response of electrons. Although massless photons enable high-speed and low-loss analog convolutions, two existing all-optical approaches including Fourier filtering and Green’s function have either limited functionality or bulky volume, thus restricting their applications in smart systems. Here, we report all-optical convolutional computing with a metasurface-singlet or -doublet imager, considered as the third approach, where its point spread function is modified arbitrarily via a complex-amplitude meta-modulator that enables functionality-unlimited kernels. Beyond one- and two-dimensional spatial differentiation, we demonstrate real-time, parallel, and analog convolutional processing of optical and biological specimens with challenging pepper-salt denoising and edge enhancement, which significantly enrich the toolkit of all-optical computing. Such meta-imager approach bridges multi-functionality and high-integration in all-optical convolutions, meanwhile possessing good architecture compatibility with digital convolutional neural networks. An ultra-compact metasurface-based imager with modified point spread function is demonstrated to realize arbitrary all-optical parallel picture convolution that is highly compatible to convolutional neural network.