Research map / Image generation / Image restoration
Geospatial Correspondences for Multimodal Registration
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 SDSN instead of color features | GT transfer with kNN classification · Zurich GT transfer, mean over 5 image pairs and both directions | average classification accuracy: 0.544 -> 0.649 | Helped |
| Combine SDSN with SIFT instead of using SIFT alone | GT transfer with kNN classification · Zurich GT transfer, mean over 5 image pairs and both directions | average classification accuracy: 0.607 -> 0.660 | Helped |
| Use SDSN+SIFT instead of affine registration | GT transfer · Zurich GT transfer | average accuracy: 63.7% -> 66.0% | Helped |
| Use SDSN+SIFT instead of non-rigid registration | GT transfer · Zurich GT transfer | average accuracy: 64.0% -> 66.0% | Helped |
| Increase the SDSN downsampling parameter from d=2 to d=20 (ablation) | SDSN feature extraction · Images 700 × 1000 pixels | time (s): 4.71 -> 0.073 | Helped |
| Increase the SDSN downsampling parameter from d=2 to d=20 (ablation) | SDSN feature extraction · GT transfer under downsampling sensitivity tests | average accuracy | No clear effect |
| Apply 42% land-cover change | SDSN matching · Synthetic land-cover change detection | recall: 84 | Hurt |
| Apply 50 m displacement | SDSN matching · Synthetic registration displacement | recall: 82 | Hurt |
| Apply 10° rotation | SDSN matching · Synthetic image rotation | recall: 57 | Hurt |
| Combine SDSN with SIFT rather than using SDSN alone | Unsupervised change detection · Zurich image pair | change detection performance | Helped |
| Use SDSN instead of color or SIFT features | Unsupervised change detection · Zurich image pair | ROC performance | Helped |
Closest papers by meaning
- Bridging Remote Sensors with Multisensor Geospatial Foundation Models (CVPR 2024 · similarity 0.85)
- GeoLink: Empowering Remote Sensing Foundation Model with OpenStreetMap Data (NeurIPS 2025 · similarity 0.85)
- Learning without Exact Guidance: Updating Large-scale High-resolution Land Cover Maps from Low-resolution Historical Labels (CVPR 2024 · similarity 0.84)
- GeoViS: Geospatially Rewarded Visual Search for Remote Sensing Visual Grounding (CVPR 2026 · similarity 0.84)
- SatCLIP: Global, General-Purpose Location Embeddings with Satellite Imagery (AAAI 2025 · similarity 0.83)
- SkyMoE: A Vision-Language Foundation Model for Enhancing Geospatial Interpretation with Mixture of Experts (AAAI 2026 · similarity 0.83)
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.