• (en anglais) Machine Learning for Enhanced Silicon Photonics Devices

    A partir de juillet 2024

    Daniele Melati - 01 70 27 06 16

    Département Photonique


    We are seeking a postdoctoral researcher with interest and experience in integrated photonics to join the Silicon Photonics team at the Center for Nanoscience and Nanotechnology (C2N). This project stems from a long lasting collaboration with the National Research Council Canada (NRC) and aims at developing accurate silicon photonics surrogate models using machine learning. In particular, we will address the key open challenge of modeling the impact of fabrication and operational environment on the perfromance of photonic devices. We will exploit both simulations and optical measurements in a data-efficient manner, and leverage transfer learning and active learning methods. The developed models will then be used to anticipate and pre-emptively compensate for fabrication errors, providing an innovative approach to design and calibrate large photonic circuits

    The successful candidate will be responsible for the identification, implementation, and training of neural network architectures both for optical behavior prediction and inverse design purposes. He/she will contribute to the design of photonic components and the experimental characterizations required to build the dataset for neural network training. Activities will be carried out in collaboration with the PI and the NRC team. Short stays in Ottawa (Canada) are foreseen as part of the project.