Research map / Vision-language and multimodal / Multimodal LLMs
GEOBench-VLM: Benchmarking Vision-Language Models for Geospatial Tasks
Muhammad Danish, Muhammad Akhtar Munir, Syed Roshaan Ali Shah, Kartik Kuckreja, Fahad Shahbaz Khan, Paolo Fraccaro, Alexandre Lacoste, Salman Khan
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 geospatial-specific EarthDial instead of GPT-4o | VLM on geospatial MCQ · Scene Understanding | accuracy: 0.7114 -> 0.7705 | Improved |
| use geospatial-specific EarthDial instead of GPT-4o | VLM on geospatial MCQ · Object Classification | accuracy: 0.5863 -> 0.4039 | Hurt |
| use geospatial-specific EarthDial instead of GPT-4o | VLM on geospatial MCQ · Event Detection | accuracy: 0.4726 -> 0.5418 | Improved |
| use geospatial-specific GeoChat instead of generic LLaVA-1.5 | VLM · Image Captioning | BERTScore: 0.6346 -> 0.4395 | Hurt |
| use generic LLaVA-OneVision instead of domain models for counting | VLM · Counting | accuracy: 0.3626 -> 0.4377 | Improved |
| use SPHINX instead of GPT-4o for referring detection | VLM · Referring Expression Detection | precision at IoU 0.5: 0.0087 -> 0.3408 | Improved |
| use multi-temporal instead of single-temporal input for land use classification | GPT-4o, EarthDial, Qwen2-VL, LLaVA-OneVision · Land Use Classification | accuracy | Improved |
| use multi-temporal instead of single-temporal input for crop classification | VLMs · Crop Classification | accuracy | Hurt |
| use multi-temporal input for disaster classification | VLMs · Disaster Classification | accuracy | Mixed |
| use multispectral instead of RGB inputs | GPT-4o and Qwen2-VL · Crop Type Classification | accuracy | Hurt |
Closest papers by meaning
- VLRS-Bench: A Vision-Language Reasoning Benchmark for Remote Sensing (NeurIPS 2026 · similarity 0.90)
- VRSBench: A Versatile Vision-Language Benchmark Dataset for Remote Sensing Image Understanding (NeurIPS 2024 · similarity 0.90)
- GeoChat: Grounded Large Vision-Language Model for Remote Sensing (CVPR 2024 · similarity 0.89)
- GTR-Bench: Evaluating Geo-Temporal Reasoning in Vision-Language Models (ICLR 2026 · similarity 0.89)
- DynamicVL: Benchmarking Multimodal Large Language Models for Dynamic City Understanding (NeurIPS 2025 · similarity 0.89)
- GeoViS: Geospatially Rewarded Visual Search for Remote Sensing Visual Grounding (CVPR 2026 · 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.
