Photonics and AI - 1Virtual room: Photons Canada - 4
|Tuesday, May 26|
AI-1-26-1 / Analysis of Two MZI-Based Topologies for Optical Neural Network
* Simon Geoffroy-Gagnon, McGill University, Canada
Farhad Shokraneh, McGill University
Odile Liboiron-Ladouceur, McGill University
Mach Zehnder Interferometer (MZI) are promising candidates for power efficient optical neural networks (ONN). This paper investigates how the ONN accuracies of two previously proposed topologies depend on the related MZI meshes and the possible imperfections in the optical devices.
AI-1-26-2 / Machine learning for quantum information processing
* Fabio Sciarrino, TBA, Canada
AI-1-26-3 / Photonic architecture for reinforcement learning
* Fulvio Flamini, University of Innsbruck, Austria
Arne Hamann, University of Innsbruck
Sofiène Jerbi, University of Innsbruck
Lea M. Trenkwalder, University of Innsbruck
Hendrik Poulsen Nautrup, University of Innsbruck
Hans J. Briegel, University of Innsbruck
Reinforcement learning algorithms are a powerful tool to manage the interaction between a system and its environment. Here we present an approach to apply these algorithms within modern-day photonic technologies. Numerical tests, performed on typical learning tasks, show that the architecture is robust against experimental noise, which can even be beneficial for the learning process. The proposed architecture, based on single-photon evolution on a tree structure of tunable beamsplitters, is simple, easy to implement and an integration in quantum optics applications appears to be within the reach of near-term technology.
AI-1-26-4 / Photonics: a great testing-ground to develop new AI algorithms for science
* Marin Soljacic, MIT, Canada
The recent AI revolution presents a number of exciting opportunities for photonics to help further advances in AI; possible applications include AI algorithms for science, but even beyond science. Some of our recent work in these topics will be presented.