TR2019-027

Nanostructured Photonic Power Splitter Design via Convolutional Neural Networks


    •  TaherSima, M., Kojima, K., Koike-Akino, T., Jha, D.K., Wang, B., Lin, C., Parsons, K., "Nanostructured Photonic Power Splitter Design via Convolutional Neural Networks", Conference on Lasers and Electro-Optics (CLEO), DOI: 10.1364/​CLEO_SI.2019.SW4J.6, May 2019.
      BibTeX TR2019-027 PDF
      • @inproceedings{TaherSima2019may,
      • author = {TaherSima, Mohammad and Kojima, Keisuke and Koike-Akino, Toshiaki and Jha, Devesh K. and Wang, Bingnan and Lin, Chungwei and Parsons, Kieran},
      • title = {Nanostructured Photonic Power Splitter Design via Convolutional Neural Networks},
      • booktitle = {Conference on Lasers and Electro-Optics (CLEO)},
      • year = 2019,
      • month = may,
      • publisher = {Optical Society of America},
      • doi = {10.1364/CLEO_SI.2019.SW4J.6},
      • url = {https://www.merl.com/publications/TR2019-027}
      • }
  • MERL Contacts:
  • Research Areas:

    Artificial Intelligence, Communications, Electronic and Photonic Devices

Abstract:

We train a convolutional neural network (CNN) that can predict the optical response of randomly generated nanopatterned photonic power splitters in a 2 to the 400th power design space with a prediction correlation coefficient of 85%.