A Segmentation Map Difference-Based Domain Adaptive Change Detection Method

Abstract

Deep Neural Network (DNN) has been widely used in remote sensing image change detection(CD) in recent years. Due to the scarcity of training data, a large number of labeled data onto other fields become the source of deep neural network concept learning in remote sensing image change detection. However, the distribution of features of the change detection data and other data varies greatly, which prevents DNN from being better applied for one task to another. To solve this problem, a domain adaptive CD method based on segmentation map difference is proposed to this paper, which includes the pre-training stage and the change detection stage. In the pre-training stage, the domain adaptive UNet (Ada-UNet) is applied as the basic network of remote sensing image segmentation for network training, with the purpose of learning the concepts of different features. In the change detection stage, strict threshold segmentation results are used to train the channel attention network, which makes it more efficient to utilize the high-dimensional feature map. The probabilistic map generated by the three-channel attention networks is evaluated, and then it is used to accurately classify the changing pixels. In this paper, experiments are carried out on datasets with different feature distributions. The results show that this method has strong domain adaptability and can greatly reduce the influence of the difference in feature distributions of the CD results.

Application-Oriented Digital Image Correlation for the High-Speed Deformation and Fracture Analysis of AISI 1045 and Ti6Al4V Materials

“…However, the frame still included the whole specimen geometry (Figure 5b). It is apparent that Pre-trained networks consistently outperform random initialized networks [26]. The decoder is pre-trained by an autoencoder to improve the segmentation quality.…”

“…The network architecture with the number of kernels used for each convolution in the encoder and decoder blocks is shown in Figure 4. Pre-trained networks consistently outperform random initialized networks [26]. The decoder is pre-trained by an autoencoder to improve the segmentation quality.…”