Research map / Vision-language and multimodal / Multimodal LLMs
Towards Interactive Global Geolocation Assistant
Zhiyang Dou, Zipeng Wang, Xumeng Han, Guorong Li, Zhenjun Han, Zhipei Huang
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 Tier-2 knowledge-augmented guidance (question plus factual prior) | GaGA · Diagnostic set (547 images) | country accuracy (%): 64.89 -> 74.77 (+9.88) | Improved |
| add Tier-2 knowledge-augmented guidance | GaGA · Diagnostic set (547 images) | city accuracy (%): 7.67 -> 9.87 (+2.20) | Improved |
| add Tier-1 focus-based inquiry | GaGA · Diagnostic set (547 images) | country accuracy (%): 64.89 -> 61.24 (-3.65) | Hurt |
| add Tier-1 focus-based inquiry | GaGA · Diagnostic set (547 images) | region and city accuracy (%): 28 -> 29.25 region (+1.28); 7.67 -> 8.22 city (+0.55) | Mixed |
| use high-specificity clues instead of ambiguous clues | GaGA · Clue specificity subsets | country accuracy (%): 59.5 -> 62.2 | Improved |
| use high-specificity clues instead of ambiguous clues | GaGA · Clue specificity subsets | city accuracy (%): 7.0 -> 9.1 | Improved |
| add Dialog data to Meta training data (ablation) | GaGA · MG-Geo data ablation | Informativeness: 0.616 -> 0.754 | Improved |
| add Dialog data to Meta training data (ablation) | GaGA · MG-Geo data ablation | Geoscore: 3872 -> 3854 | Hurt |
| add Clue data to Meta training data (ablation) | GaGA · MG-Geo data ablation | Geoscore: 3872 -> 3835 | Hurt |
| add Clue data to Meta+Dialog (ablation) | GaGA · MG-Geo data ablation | Informativeness: 0.754 -> 0.780 | Improved |
| add Introspection-tagged dialogs to Stage-2 data (ablation) | GaGA · GWS15k (reproduced) | accuracy @750km (%): 65.93 -> 78.70 (+12.77) | Improved |
| add Introspection-tagged dialogs to Stage-2 data (ablation) | GaGA · GWS15k (reproduced) | city accuracy (%): 7.96 -> 8.15 (+0.19) | No clear effect |
Closest papers by meaning
- Vision-Language Reasoning for Geolocalization: A Reinforcement Learning Approach (AAAI 2026 · similarity 0.91)
- Recognition through Reasoning: Reinforcing Image Geo-localization with Large Vision-Language Models (NeurIPS 2025 · similarity 0.90)
- GeoReasoner: Geo-localization with Reasoning in Street Views using a Large Vision-Language Model (ICML 2024 · similarity 0.90)
- G3: An Effective and Adaptive Framework for Worldwide Geolocalization Using Large Multi-Modality Models (NeurIPS 2024 · similarity 0.90)
- GeoRouter: Dynamic Paradigm Routing for Worldwide Image Geolocalization (NeurIPS 2026 · similarity 0.90)
- GRE Suite: Geo-localization Inference via Fine-Tuned Vision-Language Models and Enhanced Reasoning Chains (NeurIPS 2025 · similarity 0.89)
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.
