期刊:Frontiers in Optics [Optica Publishing Group] 日期:2023-01-01卷期号:: FM1B.3-FM1B.3
标识
DOI:10.1364/fio.2023.fm1b.3
摘要
We experimentally demonstrate a compressive sensing method by utilizing single-pixel detection at a single-photon level and deep neural network. As a benchmark test, MNIST handwritten digits are classified with 98.8% accuracy.