stanford-crfm/image2struct-latex-v1

Image2Struct - Latex

Paper | Website | Datasets (Webpages, Latex, Music sheets) | Leaderboard | HELM repo | Image2Struct repo

License: Apache License Version 2.0, January 2004

Dataset description

Image2struct is a benchmark for evaluating vision-language models in practical tasks of extracting structured information from images. This subdataset focuses on LaTeX code. The model is given an image of the expected output with the prompt: Please provide the LaTex code used to generate this image. Only generate the code relevant to what you see. Your code will be surrounded by all the imports necessary as well as the begin and end document delimiters.

The subjects were collected on ArXiv and are: eess, cs, stat, math, physics, econ, q-bio, q-fin.

The dataset is divided into 5 categories. There are 4 categories that are collected automatically using the Image2Struct repo:

• equations
• tables
• algorithms
• code

The last category: wild, was collected by taking screenshots of equations in the Wikipedia page of "equation" and its related pages.

Uses

To load the subset equation of the dataset to be sent to the model under evaluation in Python:

import datasets
datasets.load_dataset("stanford-crfm/i2s-latex", "equation", split="validation")

To evaluate a model on Image2Latex (equation) using HELM, run the following command-line commands:

pip install crfm-helm
helm-run --run-entries image2latex:subset=equation,model=vlm --models-to-run google/gemini-pro-vision --suite my-suite-i2s --max-eval-instances 10

You can also run the evaluation for only a specific subset and difficulty:

helm-run --run-entries image2latex:subset=equation,difficulty=hard,model=vlm --models-to-run google/gemini-pro-vision --suite my-suite-i2s --max-eval-instances 10

For more information on running Image2Struct using HELM, refer to the HELM documentation and the article on reproducing leaderboards.

Citation

BibTeX:

@misc{roberts2024image2struct,
      title={Image2Struct: A Benchmark for Evaluating Vision-Language Models in Extracting Structured Information from Images}, 
      author={Josselin Somerville Roberts and Tony Lee and Chi Heem Wong and Michihiro Yasunaga and Yifan Mai and Percy Liang},
      year={2024},
      eprint={TBD},
      archivePrefix={arXiv},
      primaryClass={TBD}
}