CDMamba: Remote Sensing Image Change Detection with Mamba

Haotian Zhang,Keyan Chen,Chenyang Liu,Hao Chen,Zhengxia Zou,Zhenwei Shi
2024-06-07
Abstract:Recently, the Mamba architecture based on state space models has demonstrated remarkable performance in a series of natural language processing tasks and has been rapidly applied to remote sensing change detection (CD) tasks. However, most methods enhance the global receptive field by directly modifying the scanning mode of Mamba, neglecting the crucial role that local information plays in dense prediction tasks (e.g., CD). In this article, we propose a model called CDMamba, which effectively combines global and local features for handling CD tasks. Specifically, the Scaled Residual ConvMamba (SRCM) block is proposed to utilize the ability of Mamba to extract global features and convolution to enhance the local details, to alleviate the issue that current Mamba-based methods lack detailed clues and are difficult to achieve fine detection in dense prediction tasks. Furthermore, considering the characteristics of bi-temporal feature interaction required for CD, the Adaptive Global Local Guided Fusion (AGLGF) block is proposed to dynamically facilitate the bi-temporal interaction guided by other temporal global/local features. Our intuition is that more discriminative change features can be acquired with the guidance of other temporal features. Extensive experiments on three datasets demonstrate that our proposed CDMamba outperforms the current state-of-the-art methods. Our code will be open-sourced at <a class="link-external link-https" href="https://github.com/zmoka-zht/CDMamba" rel="external noopener nofollow">this https URL</a>.
Computer Vision and Pattern Recognition
What problem does this paper attempt to address?
The paper aims to address key issues in the task of change detection (CD) in high-resolution optical remote sensing images. Specifically, the research team proposes a new model named CDMamba, which combines global and local features to handle the change detection task. The paper mainly addresses the following points: 1. **Limitations of existing methods**: Current Mamba-based methods often enhance the global receptive field by modifying the scanning pattern when dealing with dense prediction tasks (such as change detection), but they overlook the importance of local information. CDMamba overcomes this limitation by integrating both global and local information. 2. **Lack of detailed features**: The existing Mamba architecture struggles to capture detailed features in dense prediction tasks like change detection, resulting in poor fine-grained detection performance. CDMamba enhances the ability to capture local details by introducing the Scaled Residual ConvMamba (SRCM) module. 3. **Temporal feature interaction**: Considering the need for interaction between dual temporal features in change detection tasks, the paper proposes the Adaptive Global Local Guided Fusion (AGLGF) block, which dynamically promotes the interaction between dual temporal features to obtain more discriminative change features. Through the above innovations, CDMamba has been extensively tested on three different datasets, and the results show that its performance surpasses the current state-of-the-art methods. This indicates that CDMamba has significant advantages in the field of change detection.