Item

Abstract

We train convolutional neural networks to predict whether or not a set of measurements is informationally complete to uniquely reconstruct any given quantum state with no prior information. In addition, we perform fidelity benchmarking based on this measurement set without explicitly carrying out state tomography. The networks are trained to recognize the fidelity and a
reliable measure for informational completeness through collective encoding of quantum measurements, data and target states into grayscale images. By
gradually accumulating measurements and data, these convolutional networks can efficiently certify a low-measurement-cost quantum-state characterization
scheme. We confirm the potential of this machine-learning approach by presenting experimental results for both spatial-mode and multiphoton systems
of large dimensions. These predictions are further shown to improve with noise recognition when the networks are trained with additional bootstrapped training sets from real experimental data.