Research map / Image generation / Text-to-image diffusion
Scalable Multi-Temporal Remote Sensing Change Data Generation via Simulating Stochastic Change Process
Zhuo Zheng, Shiqi Tian, Ailong Ma, Liangpei Zhang, Yanfei Zhong
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 |
|---|---|---|---|
| add de-styling to the masked transition layer (ablation) | Changen · xView2 image quality | FID: 37.49 -> 34.74 | Improved |
| add de-styling to the masked transition layer (ablation) | Changen · 5% LEVIR-CD transfer | F1: 88.3 -> 88.4 | No clear effect |
| add masking to the masked transition layer (ablation) | Changen · 5% LEVIR-CD transfer | F1: 85.2 -> 88.3 | Improved |
| add masking to the masked transition layer (ablation) | Changen · xView2 image quality | FID: 20.38 -> 37.49 | Hurt |
| condition on pre-event image (Changen) versus OASIS+GPCM | Change generator · xView2 image quality | FID: 45.13 -> 34.74 | Improved |
| condition on pre-event image (Changen) versus OASIS+GPCM | Change generator · 5% LEVIR-CD transfer | F1: 86.3 -> 88.4 | Improved |
| use Changen synthetic pretraining instead of ImageNet supervised | ChangeStar R-18 · LEVIR-CD (5% train) | F1: 85.2 -> 88.4 | Improved |
| use Changen synthetic pretraining instead of ImageNet supervised | ChangeStar R-18 · LEVIR-CD (100% train) | F1: 90.5 -> 91.1 | Improved |
| pretrain on Changen-90k | ChangeStar MiT-B1 · S2Looking test | F1: 64.3 -> 67.9 | Improved |
| pretrain on Changen-90k | ChangeStar R-18 · S2Looking test | F1: 66.3 -> 67.1 | Improved |
| scale synthetic data from Changen-90k to Changen-1.4M | ChangeStar MiT-B1 · LEVIR-CD test | F1: 91.5 -> 91.7 | Improved |
| pretrain with self-supervised SeCo instead of ImageNet classification | ChangeStar R-18 · LEVIR-CD test | F1: 90.5 -> 89.9 | Hurt |
Closest papers by meaning
- Changen2: Multi-Temporal Remote Sensing Generative Change Foundation Model (TPAMI 2024 · similarity 0.96)
- The Change You Want To Detect: Semantic Change Detection In Earth Observation With Hybrid Data Generationf (CVPR 2025 · similarity 0.89)
- Feature Spectrum Learning for Remote Sensing Change Detection (CVPR 2025 · similarity 0.87)
- MMM-RS: A Multi-modal, Multi-GSD, Multi-scene Remote Sensing Dataset and Benchmark for Text-to-Image Generation (NeurIPS 2024 · similarity 0.86)
- FM-ChangeNet: Learning Change through Pathwise Feature Transport (NeurIPS 2026 · similarity 0.86)
- The STVchrono Dataset: Towards Continuous Change Recognition in Time (CVPR 2024 · similarity 0.85)
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.
