Research map / Vision-language and multimodal / Cross-modal learning
AerialFusion: Co-Motion-Driven Unified Registration and Fusion on Multi-modal Data Streams from Aerial View
Junhui Qiu, Xiang Xiang, Hongyun Wang, Jiaqi Gui
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 Skewed Motion Distribution Field (SMDF) (ablation) | AerialFusion · EUM3D fusion | EN: 5.48 -> 7.84 | Improved |
| add SMDF (ablation) | AerialFusion · EUM3D fusion | QSSIM: 0.46 -> 0.68 | Improved |
| add SMDF (ablation) | AerialFusion · EUM3D registration | MSE: 10.23 -> 9.98 | Improved |
| add registration loss Lreg (ablation) | AerialFusion · EUM3D fusion | EN: 6.99 -> 7.84 | Improved |
| add registration loss Lreg (ablation) | AerialFusion · EUM3D registration | MSE: 10.11 -> 9.98 | Improved |
| add SMDF fusion loss (ablation) | AerialFusion · EUM3D fusion | QSSIM: 0.61 -> 0.68 | Improved |
| add SMDF fusion loss (ablation) | AerialFusion · EUM3D registration | MSE: 9.99 -> 9.98 | No clear effect |
| use AerialFusion instead of RDD + LightGlue | AerialFusion · EUM3D (visible-infrared) | pose AUC@5: 61.23 -> 62.62 | Improved |
| use AerialFusion instead of RoMa on MegaDepth | AerialFusion · MegaDepth-1500 | pose AUC@5: 62.56 -> 63.08 | Improved |
| AerialFusion vs C2RF on LLVIP/VTUAV/EUM3D fusion | AerialFusion · multi-dataset fusion | EN / QAB/F | Mixed |
| AerialFusion vs Xfeat* on EUM3D registration | AerialFusion · EUM3D registration | MSE: 10.00 -> 9.98 | No clear effect |
Closest papers by meaning
- UAVScenes: A Multi-Modal Dataset for UAVs (ICCV 2025 · similarity 0.87)
- VVSim: A Large-Scale Aerial-Ground Dataset and Benchmark for Cooperative Perception (ECCV 2026 · similarity 0.86)
- EventFly: Event Camera Perception from Ground to the Sky (CVPR 2025 · similarity 0.84)
- AGC-Drive: A Large-Scale Dataset for Real-World Aerial-Ground Collaboration in Driving Scenarios (NeurIPS 2025 · similarity 0.84)
- Bring Event into RGB and LiDAR: Hierarchical Visual-Motion Fusion for Scene Flow (CVPR 2024 · similarity 0.84)
- Seeing Across Skies and Streets: Feedforward 3D Reconstruction from Satellite, Drone, and Ground Images (NeurIPS 2026 · similarity 0.84)
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.
