Research map / 3D vision / Depth and stereo
Image-Based Geo-Localization Using Satellite Imagery
Sixing Hu, Gim Hee Lee
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 |
|---|---|---|---|
| use separate (unshared) NetVLAD branches (CVM-Net-I) instead of shared weights (CVM-Net-II) (ablation) | CVM-Net · CVUSA panorama | recall@top 1%: 87.2% -> 96.3% | Improved |
| use separate NetVLAD branches instead of shared weights (ablation) | CVM-Net · Vo and Hays cropped | recall@top 1%: 66.6% -> 67.9% | Improved |
| use VGG-16 instead of ResNet-50 as local feature extractor (ablation) | CVM-Net-I · CVUSA | recall@top 1%: 88.0% -> 96.3% | Improved |
| use VGG-16 instead of DenseNet-121 (ablation) | CVM-Net-I · CVUSA | recall@top 1%: 89.0% -> 96.3% | Improved |
| use VGG-16 instead of Xception (ablation) | CVM-Net-I · CVUSA | recall@top 1%: 93.2% -> 96.3% | Improved |
| use VGG16 instead of AlexNet backbone (triplet loss) (ablation) | CVM-Net-I · CVUSA | recall@top 1%: 65.4% -> 96.3% | Improved |
| use quadruplet loss instead of triplet loss with AlexNet (ablation) | CVM-Net-I · CVUSA | recall@top 1%: 65.4% -> 73.7% | Improved |
| use quadruplet loss instead of triplet loss with AlexNet (ablation) | CVM-Net-II · CVUSA | recall@top 1%: 63.0% -> 83.9% | Improved |
| use quadruplet loss instead of triplet loss with VGG16 (ablation) | CVM-Net-I · CVUSA | recall@top 1%: 96.3% -> 89.9% | Hurt |
| use quadruplet loss instead of triplet loss with VGG16 (ablation) | CVM-Net-II · CVUSA | recall@top 1%: 87.2% -> 88.7% | No clear effect |
| use triplet loss instead of contrastive loss (ablation) | CVM-Net-I · CVUSA | recall@top 1%: 87.8% -> 96.3% | Improved |
| use triplet loss instead of contrastive loss (ablation) | CVM-Net-II · CVUSA | recall@top 1%: 79.8% -> 87.2% | Improved |
Closest papers by meaning
- CVM-Net: Cross-View Matching Network for Image-Based Ground-to-Aerial Geo-Localization (CVPR 2018 · similarity 0.91)
- Accurate 3-DoF Camera Geo-Localization via Ground-to-Satellite Image Matching (TPAMI 2022 · similarity 0.90)
- Beyond Cross-View Image Retrieval: Highly Accurate Vehicle Localization Using Satellite Image (CVPR 2022 · similarity 0.89)
- From Coarse to Fine: A Matching and Alignment Framework for Unsupervised Cross-View Geo-Localization (AAAI 2025 · similarity 0.88)
- Unleashing Unlabeled Data: A Paradigm for Cross-View Geo-Localization (CVPR 2024 · similarity 0.86)
- Game4Loc: A UAV Geo-Localization Benchmark from Game Data (AAAI 2025 · similarity 0.86)
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.
