GeoLLM-Engine: A Realistic Environment for Building Geospatial Copilots

Simranjit Singh, Michael Fore, Dimitrios Stamoulis · CVPR Workshops 2024 · 2024

DOI arXiv

Status

To read

Topics

agents

Feeds

P2 · NeurIPS 2027

What they did

Provides GeoLLM-Engine, an environment for tool-augmented geospatial copilots with geospatial API tools, dynamic maps and multimodal knowledge bases, to evaluate agents on realistic remote sensing tasks.

My note

No note yet.

BibTeX

@inproceedings{singh2024-geollm-engine,
  title = {{GeoLLM-Engine: A Realistic Environment for Building Geospatial Copilots}},
  author = {Simranjit Singh and Michael Fore and Dimitrios Stamoulis},
  booktitle = {CVPR Workshops 2024},
  year = {2024},
  doi = {10.1109/CVPRW63382.2024.00063},
  eprint = {2404.15500},
  archivePrefix = {arXiv}
}