Quantum process tomography of structured optical gates with convolutional neural networks

• DOI: 10.48550/arxiv.2402.16616
• Publication Date: 2024-02-26
• Authors: Tareq Jaouni; Francesco Di Cola...; Lorenzo Amato; Filippo Cardano; Ebrahim Karimi

The characterization of a unitary gate is experimentally accomplished via Quantum Process Tomography, which combines the outcomes different projective measurements to reconstruct underlying operator. process matrix typically extracted from maximum-likelihood estimation. Recently, optimization strategies based on evolutionary and machine-learning techniques have been proposed. Here, we investigate deep-learning approach that allows for fast accurate reconstructions space-dependent SU(2) operators, only processing minimal set measurements. We train convolutional neural network scalable U-Net architecture entire experimental images in parallel. Synthetic processes are reconstructed with average fidelity above 90%. performance our routine validated complex polarization transformations. Our further expands toolbox data-driven approaches Tomography shows promise real-time optical gates.