Q-PhotoNAS: Hybrid Quantum Neural Architecture Search Framework on Photonic Devices
Researchers have developed Q-PhotoNAS, a novel framework for designing hybrid quantum-classical neural network architectures specifically for photonic devices. This system uses a genetic algorithm to automatically search for optimal configurations, considering both classical and quantum components. When tested on image classification tasks like Digits and MNIST, Q-PhotoNAS achieved high accuracies of 99.44% and 98.78% respectively, with projected fast inference times on photonic hardware. AI
IMPACT Automated architecture search for photonic quantum systems could accelerate the development and deployment of quantum AI applications.