Research map / Label-efficient and robust learning / Zero-/few-shot and meta-learning
Training and Inference Within 1 Second – Tackle Cross-Sensor Degradation of Real-World Pansharpening with Efficient Residual Feature Tailoring
Tianyu Xin, Jin-Liang Xiao, Zeyu Xia, Shan Yin, Liang-Jian Deng
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 Tailor module (ablation) | FusionNet with patch-wise strategy · WV-3 -> WV-2 | HQNR: 0.8395 -> 0.8970 | Improved |
| add Feature Tailor module (ablation) | U2Net with patch-wise strategy · WV-3 -> WV-2 | HQNR: 0.7088 -> 0.8264 | Improved |
| add patch-wise training and inference (ablation) | U2Net with Feature Tailor · WV-3 -> WV-2 | Duration (s): 7.53 -> 0.89 | Improved |
| add patch-wise training and inference (ablation) | FusionNet with Feature Tailor · WV-3 -> WV-2 | Duration (s): 0.45 -> 0.13 | Improved |
| add patch-wise training and inference (ablation) | FusionNet with Feature Tailor · WV-3 -> WV-2 | HQNR: 0.9002 -> 0.8970 | Mixed |
| add patch-wise training and inference (ablation) | U2Net with Feature Tailor · WV-3 -> WV-2 | HQNR: 0.8308 -> 0.8264 | Hurt |
| insert Feature Tailor before channel mapper instead of after (pre-CM vs post-CM) (ablation) | ERFT on FusionNet and U2Net · WV3 (HQNR) | HQNR | Improved |
| add ERFT to pretrained FusionNet | FusionNet · QB -> GF-2 | HQNR: 0.8397 -> 0.8970 | Improved |
| add ERFT to pretrained U2Net | U2Net · QB -> GF-2 | HQNR: 0.7089 -> 0.8264 | Improved |
| add ERFT to pretrained U2Net | U2Net · WV-3 -> WV-2 | HQNR: 0.8706 -> 0.9128 | Improved |
| use ERFT-U2Net instead of zero-shot ZS-Pan | U2Net · QB -> GF-2 | HQNR: 0.8925 -> 0.8264 | Hurt |
| use ERFT-FusionNet instead of zero-shot ZS-Pan | FusionNet · QB -> GF-2 / WV-3 -> WV-2 | Duration (s): 66.74 -> 0.11 (QB to GF-2) | Improved |
Closest papers by meaning
- SWIFT:A General Sensitive Weight Identification Framework for Fast Sensor-Transfer Pansharpening (AAAI 2026 · similarity 0.90)
- PanAdapter: Two-Stage Fine-Tuning with Spatial-Spectral Priors Injecting for Pansharpening (AAAI 2025 · similarity 0.89)
- A General Adaptive Dual-level Weighting Mechanism for Remote Sensing Pansharpening (CVPR 2025 · similarity 0.87)
- Dual-Granularity Semantic Guided Sparse Routing Diffusion Model for General Pansharpening (CVPR 2025 · similarity 0.86)
- Pan-LUT: Efficient Pan-sharpening via Learnable Look-Up Tables (NeurIPS 2025 · similarity 0.85)
- Unsupervised Pan-Sharpening via Mutually Guided Detail Restoration (AAAI 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.
