Presentation
18 June 2024 A high speed fully trainable laser-based neural network
Anas Skalli, Mirko Goldmann, Nasibeh Haghighi, Marcin Gebski, Stephan Reitzenstein, Tomasz Czyszanowski, James Lott, Daniel Brunner
Author Affiliations +
Abstract
We experimentally demonstrate an autonomous, fully tunable and scalable optical neural network of 400+ parallel nodes based on a large area, multimode semiconductor laser. We implement hardware compatible, online learning strategies based on reinforcement learning and evolutionary strategies and evaluate them in terms of performance and energy cost. Our system achieves high performance and a high classification bandwidth of 15KHz for the MNIST dataset. Our approach is highly scalable both in terms of classification bandwidth and neural network size due to our device's short response time (nanosecond).
Conference Presentation
© (2024) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Anas Skalli, Mirko Goldmann, Nasibeh Haghighi, Marcin Gebski, Stephan Reitzenstein, Tomasz Czyszanowski, James Lott, and Daniel Brunner "A high speed fully trainable laser-based neural network", Proc. SPIE PC13017, Machine Learning in Photonics, PC130170R (18 June 2024); https://doi.org/10.1117/12.3022442
Advertisement
Advertisement
KEYWORDS
Artificial neural networks

Education and training

Vertical cavity surface emitting lasers

Data processing

Online learning

Photonics

Excel

Back to Top