Reducing Quantum Measurement Repetitions in Image Classification through Probability Loss

Quantum computing holds significant promise for solving complex problems more efficiently than classical computers. However, a major challenge lies in the repetitive measurement steps required to obtain reliable results from quantum states. This paper introduces the concept of probability loss to reduce the need for repeated measurements in a hybrid quantum-classical image classification model. For experiment, a hybrid model is introduced which is inspired by the Vision Transformer architecture, our model divides input images into patches and processes them as sequences.