Research map / Vision-language and multimodal / Cross-modal learning
Towards Generalized Multimodal Homography Estimation
Jinkun You, Jiaxin Cheng, Jie Zhang, Yicong Zhou
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 |
|---|---|---|---|
| train on synthetic style-rendered data only (zero-shot) (ablation) | MCNet · Train GoogleMap, test GoogleEarth | MACE: 20.518 -> 1.402 | Improved |
| train on synthetic style-rendered data only (zero-shot) (ablation) | MHN · Train GoogleMap, test PDSCOCO | MACE: 33.490 -> 4.251 | Improved |
| train on synthetic data only (zero-shot) (ablation) | DHN · Train RGB-NIR, test GoogleEarth | MACE: 18.288 -> 22.933 | Hurt |
| train on synthetic data only (zero-shot) (ablation) | DHN · Train GoogleMap, test PDSCOCO | MACE: 29.856 -> 14.173 | Improved |
| apply synthesis augmentation to GoogleMap training (ablation) | MCNet · Train GoogleMap, test GoogleEarth | MACE: 20.518 -> 4.198 | Improved |
| apply synthesis augmentation to GoogleMap training (ablation) | MCNet · Train GoogleMap, test PDSCOCO | MACE: 8.202 -> 8.489 | Hurt |
| apply synthesis augmentation to GoogleMap training (ablation) | IHN · Train GoogleMap, test GoogleEarth | MACE: 3.038 -> 2.770 | Improved |
| apply synthesis augmentation to GoogleMap training (ablation) | IHN · Train GoogleMap, test PDSCOCO | MACE: 5.352 -> 4.481 | Improved |
| apply synthesis augmentation to GoogleMap training (ablation) | MCNet · GoogleMap within-dataset | MACE: 0.261 -> 0.992 | Hurt |
| apply synthesis augmentation to GoogleMap training (ablation) | IHN · GoogleMap within-dataset | MACE: 1.025 -> 1.515 | Hurt |
| apply synthesis augmentation to RGB-NIR training (ablation) | MCNet · RGB-NIR within-dataset | MACE: 3.226 -> 3.146 | Improved |
| use CCNet instead of MCNet (zero-shot) (ablation) | CCNet · GoogleMap zero-shot | MACE: 5.093 -> 4.383 | Improved |
Closest papers by meaning
- Unsupervised Homography Estimation on Multimodal Image Pair via Alternating Optimization (NeurIPS 2024 · similarity 0.91)
- SSHNet: Unsupervised Cross-modal Homography Estimation via Problem Reformulation and Split Optimization (CVPR 2025 · similarity 0.89)
- Content-Aware Unsupervised Deep Homography Estimation and Its Extensions (TPAMI 2022 · similarity 0.89)
- MCNet: Rethinking the Core Ingredients for Accurate and Efficient Homography Estimation (CVPR 2024 · similarity 0.89)
- Adapting Dense Matching for Homography Estimation with Grid-based Acceleration (CVPR 2025 · similarity 0.89)
- Noisy One-point Homographies are Surprisingly Good (CVPR 2024 · similarity 0.87)
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.
