Research map / 3D vision / Depth and stereo
GeoDTR+: Toward Generic Cross-View Geolocalization via Geometric Disentanglement
Xiaohan Zhang, Xingyu Li, Waqas Sultani, Chen Chen, Safwan Wshah
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 CHSG (ablation) | GeoDTR+ (ConvNeXt-T, GLE v2) · CVACT->CVUSA cross-area | R@1: 42.11% -> 53.89% | Improved |
| add CHSG (ablation) | GeoDTR+ (ConvNeXt-T, GLE v2) · CVUSA same-area | R@1: 95.77% -> 95.40% | No clear effect |
| use GLE v2 instead of v1 (ResNet-34, with CHSG) (ablation) | GeoDTR+ · CVUSA->CVACT cross-area | R@1: 57.03% -> 59.72% | Improved |
| use GLE v2 instead of v1 (ResNet-34, with CHSG) (ablation) | GeoDTR+ · CVACT->CVUSA cross-area | R@1: 50.63% -> 49.09% | Hurt |
| upgrade backbone from ResNet-34 to ConvNeXt-T (with CHSG) (ablation) | GeoDTR+ · CVUSA->CVACT cross-area | R@1: 59.72% -> 61.17% | Improved |
| increase number of descriptors from 2 to 8 (ablation) | GeoDTR+ · CVUSA same-area | R@1: 93.59% -> 95.40% | Improved |
| increase number of descriptors from 2 to 8 (ablation) | GeoDTR+ · CVACT->CVUSA cross-area | R@1: 50.68% -> 53.89% | Improved |
| add CHSG instead of plain LS (ablation) | GeoDTR+ · CVUSA->CVACT cross-area | R@1: 58.62% -> 61.17% | Improved |
| double batch size from 32 to 64 with LS (ablation) | GeoDTR+ · CVUSA same-area | R@1: 95.90% -> 96.24% | Improved |
| GeoDTR+ vs GeoDTR in VIGOR cross-area | GeoDTR+ · VIGOR cross-area | R@1: 30.02% -> 36.01% | Improved |
Later papers that compared with this method
3 later accepted papers report GeoDTR as a baseline in their results tables; 3 of them report a better number on most of the results they share with it.
- MRGeo: Robust Cross-View Geo-Localization of Corrupted Images via Spatial and Channel Feature Enhancement (AAAI 2026 · better on 2 of 2 shared results)
- SinGeo: Unlock Single Model's Potential for Robust Cross-View Geo-Localization (CVPR 2026 · better on 1 of 1 shared result)
- Aligning Geometric Spatial Layout in Cross-View Geo-Localization via Feature Recombination (AAAI 2024 · better on 3 of 3 shared results)
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Cross-View Geo-Localization via Learning Disentangled Geometric Layout Correspondence (AAAI 2023 · similarity 0.94)
- Geo2: Geometry-Guided Cross-view Geo-Localization and Image Synthesis (CVPR 2026 · similarity 0.92)
- Aligning Geometric Spatial Layout in Cross-View Geo-Localization via Feature Recombination (AAAI 2024 · similarity 0.91)
- Unleashing Unlabeled Data: A Paradigm for Cross-View Geo-Localization (CVPR 2024 · similarity 0.89)
- From Coarse to Fine: A Matching and Alignment Framework for Unsupervised Cross-View Geo-Localization (AAAI 2025 · similarity 0.89)
- CVGL: Causal Learning and Geometric Topology (NeurIPS 2025 · similarity 0.88)
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.
