Semantic segmentation network integrating multi-scale feature space and semantic space

Abstract

The invention relates to a semantic segmentation network integrating multi-scale feature space and semantic space, comprising: determining a backbone network of a network coding end; VGG16 as a backbone part of the network coding end, removing a fifth pooling layer and reducing one downsampling; Designing the fusion module of feature space and semantic space of network decoder; achieving High resolution and high precision semantic segmentation by using the fusion module of multi-scale feature space and semantic space. Output Semantic Segmentation Result.

Description

technical field [0001] The invention belongs to the scene understanding technology in the fields of computer vision, pattern recognition, deep learning and artificial intelligence, and in particular relates to the technology of pixel-level semantic segmentation of the scene using a deep convolutional neural network in an image or video. Background technique [0002] Such as figure 1 As shown, in order to increase the receptive field of the deep network and reduce the amount of calculation, the backbone part of the existing deep convolutional neural network is usually down-sampled by 1 / 2 times 5 times to 1 / 32 times the input image. After multiple times of downsampling, the features of small objects and the details of objects (such as edge parts) are gradually fused by the surrounding pixel areas, and the distinguishability of small object features continues to decline. Existing representative semantic segmentation methods based on deep neural networks, such as FCN [1], direc...

AI Technical Summary

Problems solved by technology

[0011] The purpose of the present invention is to overcome the problem that the existing semantic segmentation method based on pixel-by-pixel classification does not consider the structure of similar continuous pixel regions and the structural differences of

Images