Research map / Detection and segmentation / CNN features
Beyond Attention: Convolutional Global Context for Remote Sensing Change Detection
Zhenyu Yang, Gensheng Pei, Junzhu Mao, Xinhao Cai, Tao Chen, Yazhou Yao
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use GCC instead of fixed 3x3 convolution (ablation) | ChangeGCC-B · LEVIR-CD | IoU: 81.80 -> 85.28 | Improved |
| use GCC instead of fixed 7x7 convolution (ablation) | ChangeGCC-B · LEVIR-CD | IoU: 83.24 -> 85.28 | Improved |
| add cross-scale aggregation (ablation) | ChangeGCC-B · LEVIR-CD | IoU: 84.53 -> 85.28 | Improved |
| use cross-scale aggregation instead of FPN (ablation) | ChangeGCC-B · LEVIR-CD | IoU: 84.96 -> 85.28 | Improved |
| use bidirectional cross-temporal interaction instead of early fusion (ablation) | ChangeGCC-B · LEVIR-CD | IoU: 80.55 -> 85.28 | Improved |
| use cross-temporal interaction instead of Siamese difference (ablation) | ChangeGCC-B · LEVIR-CD | IoU: 82.06 -> 85.28 | Improved |
| use cross-temporal interaction instead of Siamese concatenation (ablation) | ChangeGCC-B · LEVIR-CD | IoU: 82.44 -> 85.28 | Improved |
| use cross-temporal interaction instead of FHD (ablation) | ChangeGCC-B · LEVIR-CD | IoU: 84.93 -> 85.28 | Improved |
| use cross-temporal interaction instead of STRM (ablation) | ChangeGCC-B · LEVIR-CD | IoU: 84.72 -> 85.28 | Improved |
| add Dice loss with lambda 1.0 instead of 0 (ablation) | ChangeGCC-B · LEVIR-CD | IoU: 84.97 -> 85.28 | Improved |
| set Dice weight lambda 2.0 instead of 1.0 (ablation) | ChangeGCC-B · LEVIR-CD | IoU: 85.28 -> 84.60 | Hurt |
| scale model from Tiny to Base | ChangeGCC · LEVIR-CD | IoU: 84.85 -> 85.28 | Improved |
Closest papers by meaning
- Beyond Quadratic: Linear-Time Change Detection with RWKV (AAAI 2026 · similarity 0.90)
- RSRCC: A Remote Sensing Regional Change Comprehension Benchmark Constructed via Retrieval-Augmented Best-of-𝑁 Ranking (NeurIPS 2026 · similarity 0.89)
- Towards Generalizable Scene Change Detection (CVPR 2025 · similarity 0.87)
- Feature Spectrum Learning for Remote Sensing Change Detection (CVPR 2025 · similarity 0.87)
- The Change You Want To Detect: Semantic Change Detection In Earth Observation With Hybrid Data Generationf (CVPR 2025 · similarity 0.86)
- FM-ChangeNet: Learning Change through Pathwise Feature Transport (NeurIPS 2026 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
