中国物理B ›› 2021, Vol. 30 ›› Issue (4): 48402-.doi: 10.1088/1674-1056/abf0ff

• • 上一篇    下一篇

  

  • 收稿日期:2021-02-24 修回日期:2021-03-03 接受日期:2021-03-23 出版日期:2021-03-16 发布日期:2021-04-02

Convolutional neural network for transient grating frequency-resolved optical gating trace retrieval and its algorithm optimization

Siyuan Xu(许思源)1,2, Xiaoxian Zhu(朱孝先)2,3, Ji Wang(王佶)2,4, Yuanfeng Li(李远锋)1,2, Yitan Gao(高亦谈)2,3, Kun Zhao(赵昆)2,5,†, Jiangfeng Zhu(朱江峰)1,‡, Dacheng Zhang(张大成)1, Yunlin Chen(陈云琳)4, and Zhiyi Wei(魏志义)2,3,5   

  1. 1 School of Physics and Optoelectronic Engineering, Xidian University, Xi'an 710071 China; 2 Beijing National Laboratory for Condensed Matter Physics, Institute of Physics, Chinese Academy of Sciences, Beijing 100190, China; 3 University of Chinese Academy of Sciences, Beijing 100049, China; 4 Institute of Applied Micro-Nano Materials, School of Science, Beijing Jiaotong University, Beijing 100044, China; 5 Songshan Lake Material Laboratory, Dongguan 523808, China
  • Received:2021-02-24 Revised:2021-03-03 Accepted:2021-03-23 Online:2021-03-16 Published:2021-04-02
  • Contact: †Corresponding author. E-mail: zhaokun@iphy.ac.cn ‡Corresponding author. E-mail: jfzhu@xidian.edu.cn
  • Supported by:
    Project supported by the National Key R&D Program of China (Grant No. 2017YFB0405202) and the National Natural Science Foundation of China (Grant Nos. 61690221, 91850209, and 11774277).

RichHTML

1

PDF (PC)

50

赞

Abstract: A convolutional neural network is employed to retrieve the time-domain envelop and phase of few-cycle femtosecond pulses from transient-grating frequency-resolved optical gating (TG-FROG) traces. We use theoretically generated TG-FROG traces to complete supervised trainings of the convolutional neural networks, then use similarly generated traces not included in the training dataset to test how well the networks are trained. Accurate retrieval of such traces by the neural network is realized. In our case, we find that networks with exponential linear unit (ELU) activation function perform better than those with leaky rectified linear unit (LRELU) and scaled exponential linear unit (SELU). Finally, the issues that need to be addressed for the retrieval of experimental data by this method are discussed.

Key words: transient-grating frequency-resolved optical gating, convolutional neural network, activation function, phase retrieval algorithm

中图分类号:  (Measurements in electric variables (including voltage, current, resistance, capacitance, inductance, impedance, and admittance, etc.))

  • 84.37.+q
84.35.+i (Neural networks) 07.05.Pj (Image processing) 42.65.Hw (Phase conjugation; photorefractive and Kerr effects)