GOL

[NeurIPS 2022] Geometric order learning for rank estimation [paper]

Seon-Ho Lee, Nyeong-Ho Shin, and Chang-Su Kim

---

Dependencies

• Python 3.8
• Pytorch 1.7.1

---

Datasets

• MORPH II

• For MORPH II experiments, we follow the same fold settings in this OL repo.

• Adience

• For Adience experiments, we follow the official splits.

• [CACD]
• [UTK]

---

Usage

    $ python train.py

• Modify 'cfg.dataset' and 'cfg.setting' for training on other/custom dataset
• You may need to change 'cfg.ref_point_num' and 'cfg.margin' to obtain decent results.

---

Citation

Please cite our paper if you use this repository.

    @inproceedings{GOL2022lee,
        author    = {LEE, Seon-Ho and Shin, Nyeong-Ho and Kim, Chang-Su}, 
        title     = {Geometric Order Learning for Rank Estimation}, 
        booktitle = {Advances in Neural Information Processing Systems},
        year      = {2022}
    }

---

License

MIT License