Research map / Detection and segmentation / Open-set and few-shot detection
Free‑CD: Probabilistically Decoupled Training-Free Open-Vocabulary Change Detection with Resolution-Invariant Feature Inversion
Yongshuo Zhu, LU LI, Keyan Chen, Zhenwei Shi, ZHOU Fugen
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 feature inversion and resolution-invariant constraints to JBU (RIFI-Up) (ablation) | Free-CD · SECOND | Fscd: 21.47 -> 23.19 | Improved |
| use RIFI-Up instead of bilinear upsampling (ablation) | Free-CD · HiUCD | Fscd: 28.57 -> 30.00 | Improved |
| add feature inversion only (no resolution invariance) (ablation) | Free-CD · HRSCD | Fscd: 8.29 -> 8.18 | No clear effect |
| use transposed convolution upsampler instead of JBU (ablation) | Free-CD · SECOND | Fscd: 21.53 -> 18.72 | Hurt |
| use training-free bilinear upsampler (ablation) | Free-CD · SECOND | Fscd: 23.19 -> 21.11 | No clear effect |
| RIFI-Up on RemoteCLIP backbone vs bilinear (ablation) | Free-CD · SECOND | Fscd: 17.04 -> 18.18 | Improved |
| RIFI-Up on CLIP backbone on HRSCD (ablation) | Free-CD · HRSCD | Fscd: 4.81 -> 4.03 | Hurt |
| Free-CD framework with SegEarth-OV identifier vs DynamicEarth | Free-CD · SECOND | Fscd: 20.50 -> 23.19 | Improved |
| Free-CD framework with APE identifier vs DynamicEarth | Free-CD · SECOND | Fscd: 4.74 -> 4.88 | No clear effect |
| Free-CD (APE) vs DynamicEarth (I-M-C) on building change | Free-CD · WHU-CD | IoU: 61.1 -> 58.9 | Hurt |
| Free-CD (SegEarth-OV) vs DynamicEarth (M-C-I) on building change | Free-CD · LEVIR-CD | IoU: 36.6 -> 41.9 | Improved |
Closest papers by meaning
- The Change You Want To Detect: Semantic Change Detection In Earth Observation With Hybrid Data Generationf (CVPR 2025 · similarity 0.87)
- UniChange: Unifying Change Detection with Multimodal Large Language Model (CVPR 2026 · similarity 0.87)
- RSRCC: A Remote Sensing Regional Change Comprehension Benchmark Constructed via Retrieval-Augmented Best-of-𝑁 Ranking (NeurIPS 2026 · similarity 0.86)
- Towards Generalizable Scene Change Detection (CVPR 2025 · similarity 0.86)
- Feature Spectrum Learning for Remote Sensing Change Detection (CVPR 2025 · similarity 0.85)
- Segment Any Change (NeurIPS 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.
