search

Nicholasli1995 / EvoSkeleton

Official project website for the CVPR 2020 paper (Oral Presentation) "Cascaded deep monocular 3D human pose estimation wth evolutionary training data"
https://arxiv.org/abs/2006.07778
MIT License
333 stars 44 forks source link

eval on 3DHP #22

Closed gravitychen closed 3 years ago

gravitychen commented 3 years ago

nice code, nice paper, nice job! my question is : Can you provide the code for evaluating the 3DHP datasets,(e.x. pck3d & AUC calculation) or share other repositories that have the eval. code

Nicholasli1995 commented 3 years ago

nice code, nice paper, nice job! my question is : Can you provide the code for evaluating the 3DHP datasets,(e.x. pck3d & AUC calculation) or share other repositories that have the eval. code

Hi, the evaluation code (written in MATLAB) can be downloaded from the official website "http://gvv.mpi-inf.mpg.de/3dhp-dataset/". You can also take a look at OriNet (https://github.com/chenxuluo/OriNet-demo) about how to use the evaluation code.

gravitychen commented 3 years ago

Thank you very much!

gravitychen commented 3 years ago

dear author, I wonder whether the MPI pre-train model was released or not? I follow the link to download the pre-train model, but when I inference it, It seems to have a different result to the paper described, which is 500+ MPJPE

thank a lot(BTW can I add your WeChat to talk more details?)

Nicholasli1995 commented 3 years ago

dear author, I wonder whether the MPI pre-train model was released or not? I follow the link to download the pre-train model, but when I inference it, It seems to have a different result to the paper described, which is 500+ MPJPE

thank a lot(BTW can I add your WeChat to talk more details?)

The released pre-trained model should reproduce the results for 3DHP dataset. 500+ MPJPE is unreasonable and there may be something wrong. Please send me an email and we can discuss in more details.

gravitychen commented 3 years ago

Thank you so much