search

yeyimilk / LLM-Vision-Marine-Animals

Benchmarking Large Language Models for Image Classification of Marine Mammals
0 stars 0 forks source link

readme

Benchmarking Large Language Models for Image Classification of Marine Mammals

This dataset contains images of different marine animals that was created to evalute the abilites of LLM in classifing marine mamals.

Hierarchy class
Hierarchy class

Dataset overview

Training set Validation set Testing set
# images 1163 130 130
Min # per species 11 2 2
Max # per species 37 2 2
# Group classes 4 4 4
# Medium classes 10 10 10
# Species classes 65 65 65

Cite

If you can find our project is useful, please cite our paper

@article{qi2024benchmarking,
  title={Benchmarking Large Language Models for Image Classification of Marine Mammals},
  author={Qi, Yijiashun and Cai, Shuzhang and Zhao, Zunduo and Li, Jiaming and Lin, Yanbin and Wang, Zhiqiang},
  journal={arXiv preprint arXiv:2410.19848},
  year={2024}
}

If you are interested in evaluating the LLM's abilities in guessing image's geolocation, please refer to our another project LLMGeo.