GeoLLM-Engine: A Realistic Environment for Building Geospatial Copilots
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2024
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866911851545100288 |
|---|---|
| author | Singh, Simranjit Fore, Michael Stamoulis, Dimitrios |
| author_facet | Singh, Simranjit Fore, Michael Stamoulis, Dimitrios |
| contents | Geospatial Copilots unlock unprecedented potential for performing Earth Observation (EO) applications through natural language instructions. However, existing agents rely on overly simplified single tasks and template-based prompts, creating a disconnect with real-world scenarios. In this work, we present GeoLLM-Engine, an environment for tool-augmented agents with intricate tasks routinely executed by analysts on remote sensing platforms. We enrich our environment with geospatial API tools, dynamic maps/UIs, and external multimodal knowledge bases to properly gauge an agent's proficiency in interpreting realistic high-level natural language commands and its functional correctness in task completions. By alleviating overheads typically associated with human-in-the-loop benchmark curation, we harness our massively parallel engine across 100 GPT-4-Turbo nodes, scaling to over half a million diverse multi-tool tasks and across 1.1 million satellite images. By moving beyond traditional single-task image-caption paradigms, we investigate state-of-the-art agents and prompting techniques against long-horizon prompts. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2404_15500 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | GeoLLM-Engine: A Realistic Environment for Building Geospatial Copilots Singh, Simranjit Fore, Michael Stamoulis, Dimitrios Artificial Intelligence Computation and Language Machine Learning Geospatial Copilots unlock unprecedented potential for performing Earth Observation (EO) applications through natural language instructions. However, existing agents rely on overly simplified single tasks and template-based prompts, creating a disconnect with real-world scenarios. In this work, we present GeoLLM-Engine, an environment for tool-augmented agents with intricate tasks routinely executed by analysts on remote sensing platforms. We enrich our environment with geospatial API tools, dynamic maps/UIs, and external multimodal knowledge bases to properly gauge an agent's proficiency in interpreting realistic high-level natural language commands and its functional correctness in task completions. By alleviating overheads typically associated with human-in-the-loop benchmark curation, we harness our massively parallel engine across 100 GPT-4-Turbo nodes, scaling to over half a million diverse multi-tool tasks and across 1.1 million satellite images. By moving beyond traditional single-task image-caption paradigms, we investigate state-of-the-art agents and prompting techniques against long-horizon prompts. |
| title | GeoLLM-Engine: A Realistic Environment for Building Geospatial Copilots |
| topic | Artificial Intelligence Computation and Language Machine Learning |
| url | https://arxiv.org/abs/2404.15500 |