Metadata-Version: 2.1
Name: SGBench
Version: 0.1.2
Summary: A collection of metrics for scene graph generation
License: MIT
Author: Julian Lorenz
Author-email: julian.lorenz@uni-a.de
Requires-Python: >=3.9,<4.0
Classifier: License :: OSI Approved :: MIT License
Classifier: Programming Language :: Python :: 3
Classifier: Programming Language :: Python :: 3.9
Classifier: Programming Language :: Python :: 3.10
Classifier: Programming Language :: Python :: 3.11
Requires-Dist: imagecodecs (>=2024.1.1,<2025.0.0)
Requires-Dist: numpy (>=1.26.4,<2.0.0)
Requires-Dist: pillow (>=10.3.0,<11.0.0)
Requires-Dist: tifffile (>=2024.2.12,<2025.0.0)
Description-Content-Type: text/markdown

# SGBench: A Review and Efficient Implementation of Scene Graph Generation Metrics

Published at [CVPR 2024, Scene Graphs and Graph Representation Learning Workshop](https://sites.google.com/view/sg2rl/index).

## Installation

    pip install sgbench

## Dependencies

- [NumPy](https://numpy.org/) - For faster array operations
- [Pillow](https://pillow.readthedocs.io/en/stable/index.html) - To load ground truth PNG files
- [tifffile](https://github.com/cgohlke/tifffile) - To open TIFF files
- [imagecodecs](https://github.com/cgohlke/imagecodecs) - To support compression of TIFF files

## Citation

If you find this work useful, please consider citing our paper:

``` bibtex
@misc{lorenz2024sgbench,
      title={A Review and Efficient Implementation of Scene Graph Generation Metrics},
      author={Julian Lorenz and Robin Schön and Katja Ludwig and Rainer Lienhart},
      year={2024},
      eprint={2404.09616},
      archivePrefix={arXiv},
      primaryClass={cs.CV}
}
```

