Ultrahigh-dimensional data reconstruction deep learning method based on wavelet analysis

Abstract

The invention discloses an ultrahigh-dimensional data reconstruction deep learning method based on wavelet analysis, and the method comprises the steps: expanding high-dimensional data to different frequency domain channels through high-dimensional and high-order discrete wavelet packet transformation, and achieving the reconstruction task of the high-dimensional data in combination with a plurality of parallel neural networks; according to the method, data preprocessing is firstly carried out, then wavelet packet coefficients of different frequency band sub-domains are obtained through wavelet packet transformation, an independent network is set up and trained for the wavelet packet coefficients, and the output of the network is subjected to wavelet packet inverse transformation to reconstruct an original image. According to the method, the property that each frequency domain is independent after the high-dimensional data is subjected to wavelet transformation is utilized, and the GPU memory is utilized in parallel, so that the training process of the neural network is accelerated, and a deep learning artificial task which is originally limited by hardware computing resources becomes possible. The method is also popularized to segmentation and generation tasks. For a segmentation task, a U-net network output result is subjected to deconvolution up-sampling to obtain an original image resolution segmentation label. For a generation task, the neural network of each channel is changed into a GAN.

Description

technical field [0001] The invention relates to the technical fields of image processing and neural network, in particular to a deep learning method for ultra-high-dimensional data reconstruction based on wavelet analysis. Background technique [0002] The traditional wavelet transform is proposed to solve the loss of time domain information in Fourier transform. In the field of image processing, fast discrete wavelet transform applies a series of filters to expand image information into different independent frequency domain subbands. And expressed by wavelet coefficients. [0003] CNN is the basis of the neural network that processes the image field. The convolutional layer is the core of CNN. It extracts the detailed information of the image through a series of filters and generates a feature vector map. The pooling layer introduces invariance to CNN, and at the same time downsamples to expand the receptive field of the convolution kernel of the next layer, and the netwo...

AI Technical Summary

Problems solved by technology

[0006] In the prior art, DTI (Diffusion Tensor Imaging, Diffusion Tensor Imaging) is a method to describe the brain structure through multi-directional MRI scan data. Usually, the scan data of an individual subject is several gigabytes (GB), and the depth Learning often requires a huge sample size, and the data directly involved in training is about 100GB-200GB, but the memory capacity of the current deep learning chip is limited, such as NVIDIA Tesla V100 only supports 32GB The data is stored in the video memory of the deep learning chip for deep learning training

Images