Spatial pyramid graph convolutional network implementation method

Abstract

The invention discloses a spatial pyramid graph convolution network implementation method, which comprises the following steps: firstly, generating an incidence matrix aiming at a deep network characteristic graph, and exporting an efficient calculation method by decomposing the incidence matrix and using a multiplication combination law, so that graph convolution can be directly carried out in anoriginal characteristic space. As a lightweight network, the method breaks through the limitation that graph convolution needs to be carried out in semantic nodes in previous methods, and further improves the expression ability of the network by carrying out graph reasoning on multiple scales. According to the method, the problem of insufficient perception domain of the full convolutional networkcan be effectively solved, and the performance of the full convolutional network is remarkably improved; according to the graph convolution scheme provided by the invention, significant performance is obtained in a density prediction task in computer vision, and examples and experiments fully verify the effectiveness and potential application value of the method.

Description

technical field [0001] The invention belongs to the field of graph convolution network and deep network structure design, and in particular relates to a method for realizing a spatial pyramid graph convolution network. Background technique [0002] In recent years, network structures based on fully convolutional networks have achieved great success in computer vision tasks. The fully convolutional network is only composed of convolutional layers and pooling layers. By stacking convolutional layers, the theoretical perception domain of the network can increase as the network depth increases. But their effective receptive fields are limited, so they can only capture local information for each. Therefore, it is difficult for fully convolutional networks to capture complex contextual information. For dense prediction tasks, such as semantic segmentation, depth estimation, etc., contextual information is very critical. This problem limits the performance of fully convolutional...

AI Technical Summary

Problems solved by technology

The above methods expand the perception domain and can help the network capture remote information, but there is a problem of large amount of calculation

Images