Indoor scene semantic segmentation method based on convolutional neural network

Abstract

The invention discloses an indoor scene semantic segmentation method based on a convolutional neural network. The method comprises: in a training stage, constructing a convolutional neural network which comprises an input layer, a feature extraction layer, a feature fusion layer and an output layer; inputting an original indoor scene image into the convolutional neural network for training to obtain a corresponding semantic segmentation prediction graph; calculating a loss function value between a set formed by semantic segmentation prediction images corresponding to the original indoor sceneimages and a set formed by one-hot coded images processed by corresponding real semantic segmentation images, and obtaining a final weight vector and a bias term of a convolutional neural network classification training model; and in a test stage, inputting an indoor scene image to be semantically segmented into the convolutional neural network classification training model to obtain a predicted semantic segmentation image. According to the method, the semantic segmentation efficiency and accuracy of the indoor scene image can be improved.

Description

technical field [0001] The invention relates to a deep learning semantic segmentation method, in particular to a convolutional neural network-based indoor scene semantic segmentation method. Background technique [0002] With the continuous development of robotics, computer vision and natural language processing technologies, service robots will appear widely in our lives. The indoor space will be the main place where the service robot will work. Therefore, efficient and accurate semantic segmentation of indoor scenes is becoming one of the research hotspots in the field of computer vision. [0003] Over the past few years, fully convolutional network type architectures have shown great potential for semantic segmentation tasks and have become the architecture of choice for semantic segmentation tasks for many datasets. In indoor scene semantic segmentation tasks, color images are usually segmented using depth information as supplementary information. Generally, fully con...

AI Technical Summary

Problems solved by technology

[0004] Although the dilated convolutional structure has the advantage of preserving spatial information, it usually consumes a lot of memory during training

Since the spatial resolution of the feature map is not down-sampled during the forward propagation of the network, it needs to consume more storage and computing resources for gradient calculation

Therefore, the high memory consumption makes it difficult for the network to have a deep structure

Images