Deep learning with photonic neural cellular automata by Gordon H. Y. Li & Christian R. Leefmans & James Williams & Robert M. Gray & Midya Parto & Alireza Marandi instant download

Deep learning with photonic neural cellular automata by Gordon H. Y. Li & Christian R. Leefmans & James Williams & Robert M. Gray & Midya Parto & Alireza Marandi instant download

AbstractRapid advancements in deep learning over the past decade have fueled an insatiable demand for efficient andscalable hardware. Photonics offers a promising solution by leveraging the unique properties of light. However,conventional neural network architectures, which typically require dense programmable connections, pose severalpractical challenges for photonic realizations. To overcome these limitations, we propose and experimentallydemonstrate Photonic Neural Cellular Automata (PNCA) for photonic deep learning with sparse connectivity. PNCAharnesses the speed and interconnectivity of photonics, as well as the self-organizing nature of cellular automatathrough local interactions to achieve robust, reliable, and efficient processing. We utilize linear light interference andparametric nonlinear optics for all-optical computations in a time-multiplexed photonic network to experimentallyperform self-organized image classification. We demonstrate binary (two-class) classification of images using as few as3 programmable photonic parameters, achieving high experimental accuracy with the ability to also recognize out-ofdistribution data. The proposed PNCA approach can be adapted to a wide range of existing photonic hardware andprovides a compelling alternative to conventional photonic neural networks by maximizing the advantages of light1234567890():,; 1234567890():,;1234567890():,;1234567890():,;based computing whilst mitigating their practical challenges. Our results showcase the potential of PNCA in advancingphotonic deep learning and highlights a path for next-generation photonic computers.