Image classification method and related device

Abstract

The invention discloses an image classification method, and the method comprises the steps: carrying out the quantum state processing of to-be-recognized image data, and obtaining quantum state inputdata; calculating the quantum state input data by adopting a convolution kernel provided with a quantum revolving door according to the convolution network parameters and the training parameters to obtain hidden layer data; calculating the hidden layer data by adopting a convolution kernel provided with a quantum revolving door according to the output channel parameters and the training parametersto obtain output quantum bits; and classifying the output quantum bits to obtain a classification result. The hidden layer and the output layer are calculated by adopting the convolution kernel provided with the quantum revolving door, so that the linear output of the quantum convolution network is maintained, the unknown problem is avoided, and the accuracy and precision of image recognition areimproved. The invention further discloses an image classification device, a server and a computer readable storage medium, which have the above beneficial effects.

Description

technical field [0001] The present application relates to the technical field of quantum computing, and in particular to an image classification method, an image classification device, a server, and a computer-readable storage medium. Background technique [0002] With the continuous development of quantum technology, it is applied in more and more fields. Among them, quantum information is a new discipline developed by the combination of quantum physics and information technology, mainly including two fields of quantum communication and quantum computing. Quantum communication mainly studies quantum cryptography, quantum teleportation, long-distance quantum communication technology, etc.; quantum computing mainly studies quantum computers and quantum algorithms suitable for quantum computers. [0003] In the related technologies of quantum computing, only some adjacent two-pixel pooling methods are used in the hidden layer, that is, the multiplication of two adjacent pixel...

AI Technical Summary

Problems solved by technology

However, in actual use, the quantum convolutional neural network of this classification function has the following problems: first, the size of the convolution kernel and the convolution method are fixed, and can only be applied to data with low dimensions. For the expansion of different data sets Poor performance, limited application scenarios

Then there is the nonlinear problem of the classification function, which leads to some unpredictable problems in the nonlinear calculation process, which reduces the accuracy and precision of the quantum convolution network for calculation, and further reduces the accuracy and precision of the image recognition process.

Images