Evaluation code for human pose estimation includes strict PCP, PDJ, and PCK.
Branch: master
Clone or download
Latest commit f6c71b6 Sep 10, 2016
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
gt Add evaluation protocal (PCP) for MPII Cooking Activities Dataset Sep 10, 2016
results
utils clean up Sep 10, 2016
.gitignore
LICENSE Initial commit Nov 20, 2015
README.md update readme Apr 13, 2016
demo_eval_flic.m modification of eval_pcp for better generalization Sep 10, 2016
demo_eval_flic_pc.m Add and modify the eval_pck to evaluate FLIC (PC) with original 11 ke… Sep 10, 2016
demo_eval_lsp.m modification of eval_pcp for better generalization Sep 10, 2016
demo_eval_mpiicooking.m Add evaluation protocal (PCP) for MPII Cooking Activities Dataset Sep 10, 2016
demo_eval_parse.m modification of eval_pcp for better generalization Sep 10, 2016
startup.m init Nov 23, 2015

README.md

End-to-End Learning of Deformable Mixture of Parts and Deep Convolutional Neural Networks for Human Pose Estimation

Pose estimation results on the LSP [1] dataset, the FLIC [2] dataset, and the Image Parse [3] dataset for the following paper.

Wei Yang, Wanli Ouyang, Hongsheng Li, Xiaogang Wang. "End-to-End Learning of Deformable Mixture of Parts and Deep Convolutional Neural Networks for Human Pose Estimation". In CVPR, 2016.

Instruction

Please run demo_eval_DATASETNAME.m to evaluate a specific dataset. DATASETNAME can be lsp, flic or parse.

Acknowledgement

The evaluation code for the PCP and the PCK measurements are from a widely used version from the MPII Human Pose Dataset. The code for the PDJ measurement is from Chen and Yuille, NIPS'14

Citation

@InProceedings{yang2016end,
  Title 		= {End-to-End Learning of Deformable Mixture of Parts and Deep Convolutional Neural Networks for Human Pose Estimation},
  Author 		= {Yang, Wei and Ouyang, Wanli and Li, Hongsheng and Wang, Xiaogang},
  Booktitle 	= {CVPR},
  Year 			= {2016}
}

References

  1. S. Johnson and M. Everingham. Clustered pose and nonlinear appearance models for human pose estimation. In BMVC, 2010.
  2. B. Sapp and B. Taskar. Modec: Multimodal decomposable models for human pose estimation. In CVPR, 2013.
  3. D. Ramanan. Learning to parse images of articulated objects. In NIPS, 2006.