QSurfNet: a hybrid quantum convolutional neural network for surface defect recognition

Quantum Convolutional Neural Networks for Multi-Channel Supervised Learning

“…The advances in both quantum computing and ML inspire the development of quantum ML (QML) methods to exploit the speed of quantum computations and the predictive capabilities of ML. There has been recent work demonstrating the feasibility and advantages of substituting components of classical ML architectures with quantum analogues [36][37][38][39], such as quantum circuits in place of classical convolutional kernels in convolutional neural networks (CNNs) [40][41][42][43][44][45][46][47][48]. Classical CNNs are state-of-the-art for image, video, and sound recognition tasks [49,50] and also have applications in the natural sciences [35,[51][52][53][54][55][56].…”

“…However, extension of this approach to tasks involving data with more channels is prohibited by current hardware limitations. As an attempt to overcome this challenge, there has been much recent work that performs a measurement on each channel individually, collapsing the wavefunction after measuring a given channel of the data and storing the measurement classically [42][43][44]46]. Although the hardware requirements have no dependence on the number of channels when using this method, much of the inter-channel information is lost, which is valuable for accurately modeling the data.…”