Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2504.13026 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866917988732502016 |
|---|---|
| author | Liu, Yide Sun, Haijiang Zhang, Xiaowen Liu, Qiaoyuan Chen, Zhouchang Xiao, Chongzhuo |
| author_facet | Liu, Yide Sun, Haijiang Zhang, Xiaowen Liu, Qiaoyuan Chen, Zhouchang Xiao, Chongzhuo |
| contents | Remote Sensing Image Super-Resolution (RSISR) reconstructs high-resolution (HR) remote sensing images from low-resolution inputs to support fine-grained ground object interpretation. Existing methods face three key challenges: (1) Difficulty in extracting multi-scale features from spatially heterogeneous RS scenes, (2) Limited prior information causing semantic inconsistency in reconstructions, and (3) Trade-off imbalance between geometric accuracy and visual quality. To address these issues, we propose the Texture Transfer Residual Denoising Dual Diffusion Model (TTRD3) with three innovations: First, a Multi-scale Feature Aggregation Block (MFAB) employing parallel heterogeneous convolutional kernels for multi-scale feature extraction. Second, a Sparse Texture Transfer Guidance (STTG) module that transfers HR texture priors from reference images of similar scenes. Third, a Residual Denoising Dual Diffusion Model (RDDM) framework combining residual diffusion for deterministic reconstruction and noise diffusion for diverse generation. Experiments on multi-source RS datasets demonstrate TTRD3's superiority over state-of-the-art methods, achieving 1.43% LPIPS improvement and 3.67% FID enhancement compared to best-performing baselines. Code/model: https://github.com/LED-666/TTRD3. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2504_13026 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | TTRD3: Texture Transfer Residual Denoising Dual Diffusion Model for Remote Sensing Image Super-Resolution Liu, Yide Sun, Haijiang Zhang, Xiaowen Liu, Qiaoyuan Chen, Zhouchang Xiao, Chongzhuo Computer Vision and Pattern Recognition Remote Sensing Image Super-Resolution (RSISR) reconstructs high-resolution (HR) remote sensing images from low-resolution inputs to support fine-grained ground object interpretation. Existing methods face three key challenges: (1) Difficulty in extracting multi-scale features from spatially heterogeneous RS scenes, (2) Limited prior information causing semantic inconsistency in reconstructions, and (3) Trade-off imbalance between geometric accuracy and visual quality. To address these issues, we propose the Texture Transfer Residual Denoising Dual Diffusion Model (TTRD3) with three innovations: First, a Multi-scale Feature Aggregation Block (MFAB) employing parallel heterogeneous convolutional kernels for multi-scale feature extraction. Second, a Sparse Texture Transfer Guidance (STTG) module that transfers HR texture priors from reference images of similar scenes. Third, a Residual Denoising Dual Diffusion Model (RDDM) framework combining residual diffusion for deterministic reconstruction and noise diffusion for diverse generation. Experiments on multi-source RS datasets demonstrate TTRD3's superiority over state-of-the-art methods, achieving 1.43% LPIPS improvement and 3.67% FID enhancement compared to best-performing baselines. Code/model: https://github.com/LED-666/TTRD3. |
| title | TTRD3: Texture Transfer Residual Denoising Dual Diffusion Model for Remote Sensing Image Super-Resolution |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2504.13026 |