Optical Circuit for Quantum Similarity Measurement Algorithm

We report a similarity measurement algorithm, based on quantum optics formulation, where a coherent state is used as the elementary quantum resource. The algorithm is demonstrated for an image database and we provide an architecture with constituent optical elements in linear order with respect to the image resolution. The obtained phase-distributed multimode coherent state is fed into an image retrieval scheme and we identify the appropriate laser intensity parameter for similarity measurement. We demonstrate the viability of the protocol through an objective quality assessment of images by adding consecutive layers of noises. The results are in good agreement with the expected outcome. The image distortion-sensitivity analysis of the metric establishes the further merit of the model. Our quantum algorithm has wider applicability also in supervised machine learning tasks.