Research map / Vision-language and multimodal / Cross-modal learning
Any2Any: Unified Arbitrary Modality Translation for Remote Sensing
Haoyang Chen, Jing Zhang, Di Wang, Hebaixu Wang, Shiqin Wang, Pohsun Huang, Jiayuan Li, Haonan Guo, Zheng Wang, Bo Du
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 target-specific residual adapter (ablation) | Any2Any (SAR->RGB only) · RST-1M test, SAR->RGB | PSNR: 20.68 -> 20.88 | Improved |
| add target-specific residual adapter (ablation) | Any2Any (SAR->RGB only) · RST-1M test, SAR->RGB | RMSE: 28.51 -> 27.89 | Improved |
| continue from pretrained model instead of training from scratch when adding RGB->SAR (ablation) | Any2Any · RST-1M test, SAR->RGB | PSNR: 19.63 -> 21.44 | Improved |
| continue from pretrained model instead of training from scratch when adding RGB->SAR (ablation) | Any2Any · RST-1M test, SAR->RGB | RMSE: 32.83 -> 25.87 | Improved |
| add all SAR->other translation directions to training (ablation) | Any2Any · RST-1M test, SAR->RGB | PSNR: 20.88 -> 22.06 | Improved |
| add all other->RGB translation directions to training (ablation) | Any2Any · RST-1M test, SAR->RGB | PSNR: 20.88 -> 21.36 | Improved |
| train on all 14 directions (Any->Any) (ablation) | Any2Any · RST-1M test, SAR->RGB | PSNR: 20.88 -> 22.25 | Improved |
| one unified model instead of pairwise models (ablation) | Any2Any-S · RST-1M test, SAR->RGB | PSNR: 20.81 -> 22.25 | Improved |
| one unified model instead of pairwise models (ablation) | Any2Any-S · RST-1M test, RGB->PAN | PSNR: 32.46 -> 31.30 | Improved |
| unified model instead of pairwise models (ablation) | Any2Any-S · RST-1M test, NIR->RGB | PSNR: 23.01 -> 23.01 | No clear effect |
| scale DiT backbone from S to L | Any2Any · RST-1M test, SAR->RGB | PSNR: 22.25 -> 25.20 | Improved |
| use unified Any2Any-L instead of best pairwise baseline | Any2Any-L · RST-1M test, 14 translation tasks | PSNR | Improved |
Closest papers by meaning
- MMM-RS: A Multi-modal, Multi-GSD, Multi-scene Remote Sensing Dataset and Benchmark for Text-to-Image Generation (NeurIPS 2024 · similarity 0.88)
- GeoLLaVA-8K: Scaling Remote-Sensing Multimodal Large Language Models to 8K Resolution (NeurIPS 2025 · similarity 0.87)
- SkyScript: A Large and Semantically Diverse Vision-Language Dataset for Remote Sensing (AAAI 2024 · similarity 0.87)
- SkySense: A Multi-Modal Remote Sensing Foundation Model Towards Universal Interpretation for Earth Observation Imagery (CVPR 2024 · similarity 0.87)
- GeoCoT: Towards Reliable Remote Sensing Reasoning with Manifold Perspective (CVPR 2026 · similarity 0.87)
- AnySat: One Earth Observation Model for Many Resolutions, Scales, and Modalities (CVPR 2025 · 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.
