Awesome

LearnableGroups-Hand

The code for the paper Exploiting Learnable Joint Groups for Hand Pose Estimation (Accepted by AAAI2021).

Paper

Overall network:

Qualitative Results

some qualitative results on the RHD/STB/FHD dtasets. In each triplet, from left to right: imgs (input), predictions, GT.

• RHD: you can obtain this dataset via hand3d.

• FHD: you can obtain this dataset following this instruction FreiHand .

• STB: you can obtain this dataset via STB .

Citing LearnableGroups-Hand

If this repository is helpful to your research, please cite the paper:

@misc{li2020exploiting,
      title={Exploiting Learnable Joint Groups for Hand Pose Estimation}, 
      author={Moran Li and Yuan Gao and Nong Sang},
      year={2020},
      eprint={2012.09496},
      archivePrefix={arXiv},
      primaryClass={cs.CV}
}

Usage

The code is built on Python3 and Pytorch 1.6.0.

Install dependencies

pip install -r requirements.txt

Run the code

• evaluate on the RHD:

python eval_RHD.py --data_dir 'your RHD_published_v2 dataset path'

Comparison with SOTA methods

• Plot AUC curve on RHD/STB/DO

  • obtain AUC curve for comparison with other SOTA methods (as shown in Fig.3 in main paper).
  <p align="middle"> <img src="assets/overall_auc.png", width="780"> </p>

• Ours User Name on the FreiHand CodaLab website is 'anonymous15'