In plain words: A light-based chip performs the image-filtering step that lets a neural network recognize objects, replacing the usual electronic circuits. It is designed to use only a fraction of the energy per image checked compared with today's best electronic chips.

Abstract

Convolutional Neural Networks (CNNs) are a class of Artificial Neural Networks(ANNs) that employ the method of convolving input images with filter-kernels for object recognition and classification purposes. In this paper, we propose a photonics circuit architecture which could consume a fraction of energy per inference compared with state of the art electronics.

Hengameh Bagherian, Scott Skirlo, Yichen Shen, Huaiyu Meng, Vladimir Ceperic, Marin Soljacic
arXiv:1808.03303 · cs.ET · submitted Aug 9, 2018 · updated Aug 16, 2018
abstract · pdf · html · 18 pages, 7 figures