The ARAS-Farabi Framework is designed to facilitate research in the area of skill assessment for Capsulorhexis surgery. This framework utilizes deep learning techniques for real-time performance in detecting and tracking the capsulorhexis cystotome and pupil.

To download the AFCID Dataset and related .pt files, please contact: [email protected]

Here is a video demonstration of the framework's real-time performance using the AFCID dataset.

• docs/: Documentation and guides
• src/: Source code for the framework
• tests/: Test scripts and validation tools
• data/: Sample data and scripts for handling dataset
• models/: Pre-trained models and training scripts

If you use this framework or the AFCID dataset in your research, please cite:

@INPROCEEDINGS{9663494,
  author={Ahmadi, Mohammad Javad and Allahkaram, Mohammad Sina and Rashvand, Ashkan and Lotfi, Faraz and Abdi, Parisa and Motaharifar, Mohammad and Mohammadi, S. Farzad and Taghirad, Hamid D.},
  booktitle={2021 9th RSI International Conference on Robotics and Mechatronics (ICRoM)},
  title={ARAS-Farabi Experimental Framework for Skill Assessment in Capsulorhexis Surgery},
  year={2021},
  pages={385-390},
  doi={10.1109/ICRoM54204.2021.9663494}
}