Skip to content

microsoft/lmm-graphical-perception

Repository files navigation

Evaluating Graphical Perception of Large Multimodal Models

Source code and data for the Graphical Perception Evaluation and Study: Zhang et al. 2025

Chart Reasoning (RQ1 and RQ2 in the paper)

To get started, copy openai-keys.env.template into a file named openai-keys.env and fill necessary environment variables (supports only AzureOpenAI keys at the moment). Azure OpenAI keys will be used by agents/client.py to initialize models.

Env Setup

conda create --prefix ./env/graphical_perception  python=3.9 -y

conda activate ./env/graphical_perception

pip install -r requirements.txt

Vis-text Data Downloading

cd data
chmod 777 download_data.sh
./download_data.sh --scenegraphs --vl_spec

OpenAI Key Setup

cp openai-keys.env.template openai-keys.env

then edit openai-keys.env with your own keys.

Run Chart Reasoning Pipeline

In labs/, run chart-reasoning-pipeline.ipynb to generate tasks, run GPT-4o/4v (determined by chart_reasoning/utils/clients.py), and get results evaluated by GPT-4o/4v and the detailed metrics. Note that table generation needs executable Chrome browser. To evaluate and get the final metrics for open-source models, please run chart-reasoning-oss-pipeline.ipynb

Chart Probing (RQ3 in the paper)

Please refer to chart_probing/README.md for more details.

Questions or Bugs?

If you have any questions, please feel free to contact drogozhang[AT]gmail[DOT]com or open an issue so we can help you better and quicker :)

Contributing

This project welcomes contributions and suggestions. Most contributions require you to agree to a Contributor License Agreement (CLA) declaring that you have the right to, and actually do, grant us the rights to use your contribution. For details, visit https://cla.opensource.microsoft.com.

When you submit a pull request, a CLA bot will automatically determine whether you need to provide a CLA and decorate the PR appropriately (e.g., status check, comment). Simply follow the instructions provided by the bot. You will only need to do this once across all repos using our CLA.

This project has adopted the Microsoft Open Source Code of Conduct. For more information see the Code of Conduct FAQ or contact [email protected] with any additional questions or comments.

Trademarks

This project may contain trademarks or logos for projects, products, or services. Authorized use of Microsoft trademarks or logos is subject to and must follow Microsoft's Trademark & Brand Guidelines. Use of Microsoft trademarks or logos in modified versions of this project must not cause confusion or imply Microsoft sponsorship. Any use of third-party trademarks or logos are subject to those third-party's policies.

About

Evaluating Graphical Perception of Large Multimodal Models

Resources

License

Code of conduct

Security policy

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 3

  •  
  •  
  •