The test results are of the wrong order of magnitude

zczcwh / PoseFormer

The project is an official implementation of our paper "3D Human Pose Estimation with Spatial and Temporal Transformers".

495 stars 73 forks source link

The test results are of the wrong order of magnitude #31

Open JihuaPeng opened 2 years ago

JihuaPeng commented 2 years ago

When i run the command "python run_poseformer.py -k cpn_ft_h36m_dbb -f 81 -c checkpoint --evaluate detected81f.bin" to evaluate the pre-trained 81-frame model, it produces the following result

Is there a bug in this code please?

TomCatOnline commented 2 years ago

I get the same result,do you figured out how,please?

mdimtiazh commented 1 year ago

Got the same output !!! Do you know how to deal with this? Thanks in advance !!