Change Guiding Network: Incorporating Change Prior to Guide Change Detection in Remote Sensing Imagery
The rapid advancement of automated artificial intelligence algorithms and remote sensing instruments has benefited change detection (CD) tasks. However, there is still a lot of space to study for precise detection, especially the edge integrity and internal holes phenomenon of change features. In order to solve these problems, we design the Change Guiding Network (CGNet), to tackle the insufficient expression problem of change features in the conventional U-Net structure adopted in previous methods, which causes inaccurate edge detection and internal holes. Change maps from deep features with rich semantic information are generated and used as prior information to guide multi-scale feature fusion, which can improve the expression ability of change features. Meanwhile, we propose a self-attention module named Change Guide Module (CGM), which can effectively capture the long-distance dependency among pixels and effectively overcome the problem of the insufficient receptive field of traditional convolutional neural networks. On four major CD datasets, we verify the usefulness and efficiency of the CGNet, and a large number of experiments and ablation studies demonstrate the effectiveness of CGNet. We're going to open-source our code at https://github.com/ChengxiHAN/CGNet-CD.
PDF AbstractCode
S2Looking
| Task | Dataset | Model | Metric Name | Metric Value | Global Rank | Benchmark |
|---|---|---|---|---|---|---|
| Change Detection | CDD Dataset (season-varying) | CGNet | F1-Score | 94.73 | # 13 | |
| Precision | 93.67 | # 4 | ||||
| F1 | 94.73 | # 4 | ||||
| Overall Accuracy | 98.74 | # 5 | ||||
| IoU | 90.00 | # 5 | ||||
| Recall | 95.82 | # 4 | ||||
| KC | 94.02 | # 3 | ||||
| Change Detection | DSIFN-CD | CGNet | F1 | 60.19 | # 6 | |
| IoU | 43.05 | # 5 | ||||
| Overall Accuracy | 81.71 | # 5 | ||||
| Precision | 47.75 | # 3 | ||||
| Recall | 81.38 | # 2 | ||||
| KC | 49.34 | # 3 | ||||
| Change Detection | GoogleGZ-CD | CGNet | F1 | 85.89 | # 3 | |
| Precision | 88.07 | # 1 | ||||
| Recall | 83.82 | # 3 | ||||
| Overal Accuracy | 93.23 | # 2 | ||||
| KC | 81.45 | # 2 | ||||
| IoU | 75.27 | # 2 | ||||
| Change Detection | LEVIR+ | CGNet | F1 | 83.68 | # 1 | |
| Prcision | 81.46 | # 2 | ||||
| Recall | 86.02 | # 1 | ||||
| OA | 98.63 | # 1 | ||||
| KC | 82.97 | # 1 | ||||
| IoU | 71.94 | # 1 | ||||
| Change Detection | LEVIR-CD | CGNet | F1 | 92.01 | # 7 | |
| IoU | 85.21 | # 5 | ||||
| Overall Accuracy | 99.20 | # 1 | ||||
| F1-score | 92.01 | # 2 | ||||
| Precision | 93.15 | # 2 | ||||
| Recall | 90.90 | # 3 | ||||
| Change Detection | S2Looking | CGNet | F1-Score | 64.33 | # 4 | |
| Precision | 70.18 | # 3 | ||||
| Recall | 59.38 | # 1 | ||||
| OA | 99.20 | # 3 | ||||
| KC | 63.93 | # 1 | ||||
| IoU | 47.41 | # 1 | ||||
| Change Detection | SYSU-CD | CGNet | F1 | 79.92 | # 4 | |
| Precision | 86.37 | # 1 | ||||
| Recall | 74.37 | # 3 | ||||
| OA | 91.19 | # 1 | ||||
| KC | 74.31 | # 1 | ||||
| IoU | 66.55 | # 1 | ||||
| Change Detection | WHU-CD | CGNet | F1 | 92.59 | # 3 | |
| Overall Accuracy | 99.48 | # 2 | ||||
| Precision | 94.47 | # 2 | ||||
| Recall | 90.79 | # 2 | ||||
| KC | 92.33 | # 2 | ||||
| IoU | 86.21 | # 2 |
