Semi-supervised Human Pose Estimation in Art-historical Images

This is the official GitHub page for the paper:

Matthias Springstein, Stefanie Schneider, Christian Althaus and Ralph Ewerth: "Semi-supervised Human Pose Estimation in Art-historical Images": Proceedings of the 30th ACM International Conference on Multimedia. Lisbon, 2022.

Paper

• Arxiv https://arxiv.org/abs/2207.02976
• ACM https://dl.acm.org/doi/10.1145/3503161.3548371

Slides

The slides for the presentation for ACM MM 22 are available at the following link: https://tibhannover.github.io/iart-semi-pose

Notebook

If you are interested in using this model for your own project, there is a Colab notebook that shows how to load and use the trained models.

https://githubtocolab.com/TIBHannover/iart-semi-pose/blob/main/notebooks/example.ipynb

Model

The model files are available on huggingface.

https://huggingface.co/springsteinm/iart-semi-pose