Jeonghwan Kim

Why does the test and evaluation shows different results?

3

I'm training ROMP model with your code It saves every validation process but, I test the results with "python -m romp.test" code I got the different result as evaluation I...

To reproduce the final model

6

In the paper, it says that fine tuned model shows the best performance at 120Epoch, but after running the v1_hrnet_3dpw_ft.yml file, it showed the best performance at 1 epoch. Did...

Trying to reproduce Uformer_B model for deblurring

5

I trained the model you provided on the GoPro dataset. I trained 3000 Epoch with CharbonnierLoss and Edge loss as MPRNet code and I also use same scheduler as MPRNet...

I'm trying to generate posmap from the output of 3D reconstruction model like ROMP, PARE. Is is possible to generate posmap with only vertices information? If is possible can help...

Why I got the wrong global orient results

5

 I rendered the obj with pytorch3d library and fit the SMPL skinned model using the parameters provided. but, I got the results as above. I think this is upside...

I runned demo.lua and demo.m with my own dataset, but I want to fit on smpl model which makes more detail how can I do? I think it needs info.m...

how can I run without display

2

Hi thank you for your great work! Im trying to run fit_smplicit.py at remote server(Linux) but it has an error because of the display Like this. ['data/images/back2.jpg', 'data/images/front2.jpg', 'data/images/v2.jpg'] /usr/local/lib/python3.7/dist-packages/SMPLicit-0.0.1-py3.7.egg/SMPLicit/SMPL.py:94:...