Hongwen Zhang
Hongwen Zhang
Adding `from __future__ import absolute_import` at head of train_net_step.py could solve this problem.
Hi, thanks for your interest in PyMAF-X. `projection(dict(cam_sxy=smplx_verts), pred_cam)` seems to be not correct. Do you mean `projection(smplx_verts, dict(cam_sxy=pred_cam))`? Moreover, the input resolution of PyMAF-X should be 224x224.
The demo code of PyMAF/PyMAF-X first detects the person in the image, then crops a new image based on the detected person bounding box. So, I think the reprojected points...
Yeah, the reprejection result is now correct. Glad to hear that you have figured it out! @tommycwh
Hi, what do you mean by bigger images? More than one person in the image with higher resolutions? If so, you might need a person detector to crop person regions...
In this case, you need to re-train the model with higher resolutions of input images and IUV maps. When using IUV maps with higher resolutions, the reconstruction performance should be...
Our model is trained on a single TITAN Xp GPU, which takes about one day. Feel free to let me know if you encounter problems or have any results.
> Hello. I would also like to run in higher resolution. I need to change the focal length. Do I also have to retrain it? Hi, retraining the model should...
Hi EnricoBeltramo, thanks for your interest in our work! The provided checkpoint file is trained with the male SMPL model, so the reconstruction results may be less satisfactory when different...
Duplicate of #1