计算机工程与应用 ›› 2023, Vol. 59 ›› Issue (8): 254-262.DOI: 10.3778/j.issn.1002-8331.2112-0484

• 图形图像处理 • 上一篇    下一篇

融合注意力机制的对抗式半监督语义分割

云飞,殷雁君,张文轩,智敏   

  1. 内蒙古师范大学 计算机科学技术学院,呼和浩特 010022
  • 出版日期:2023-04-15 发布日期:2023-04-15

Adversarial Semi-Supervised Semantic Segmentation with Attention Mechanism

YUN Fei, YIN Yanjun, ZHANG Wenxuan, ZHI Min   

  1. School of Computer Science and Technology, Inner Mongolia Normal University, Hohhot 010022, China
  • Online:2023-04-15 Published:2023-04-15

摘要: 图像语义分割任务是计算机视觉领域重要研究课题之一。当前基于全卷积神经网络的语义分割算法存在像素之间缺乏关联性、卷积核感受野小于理论值、人工标记数据集标签成本大等问题。为了解决上述问题,提出了一种融合注意力机制的对抗式半监督语义分割模型。将生成对抗网络应用到图像语义分割中,增强像素点之间的关联性;提出模型在生成网络中加入自注意力模块和多核池化模块以对长距离语义信息进行融合,扩大了卷积核感受野;在PASCAL VOC2012增强数据集和Cityscapes数据集上进行了大量实验,实验结果证明了该方法在图像语义分割任务中的有效性和可靠性。

关键词: 语义分割, 生成对抗网络, 注意力机制, 半监督训练

Abstract: Image semantic segmentation is one of the most important research topics in computer vision. The current semantic segmentation algorithm based on full convolutional neural network has some problems, such as lack of correlation between pixels, convolution kernel receptive field smaller than the theoretical value, and high label cost of manually labeled data set. In order to solve the above problems, an antithesis semi-supervised semantic segmentation model integrating attention mechanism is proposed. The generative adversarial network is applied to image semantic segmentation to enhance the correlation between pixels. In this model, self-attention module and multi-core pooling module are added to generate network to fuse long distance semantic information, and the convolution kernel receptive field is enlarged. A large number of experiments are carried out on PASCAL VOC2012 enhanced dataset and Cityscapes dataset, and the experimental results prove the validity and reliability of the proposed method for image semantic segmentation.

Key words: semantic segmentation, generative adversarial network, attentional mechanism, semi-supervised training