SPIN icon indicating copy to clipboard operation
SPIN copied to clipboard

Repository for the paper "Learning to Reconstruct 3D Human Pose and Shape via Model-fitting in the Loop"

Results 48 SPIN issues
Sort by recently updated
recently updated
newest added

Running `python3 demo.py --checkpoint=data/model_checkpoint.pt --img=examples/im1010.jpg --bbox=examples/im1010_bbox.json` in a docker environment using the provided Docker image results in script freezing after preprocessing the image, at the [model prediction step](https://github.com/nkolot/SPIN/blob/master/demo.py#L121). The container...

Great job! Now I get access of h36m Mosh data, but i found it wrong rendering vertices to images in training. It seems that the ''global_orient'' is wrong. Do you...

I notice that in the code, only 14 joints used for eval on h36m, why not 17 joints? `# Joint selectors` `# Indices to get the 14 LSP joints from...

Hellow,I'm trying to replicate your work right now, but I still got large error after ten epoches of training, and here are my results, I want to ask if this...

Bumps [scipy](https://github.com/scipy/scipy) from 1.0.0 to 1.10.0. Release notes Sourced from scipy's releases. SciPy 1.10.0 Release Notes SciPy 1.10.0 is the culmination of 6 months of hard work. It contains many...

dependencies

While installing packages in my Python 3 conda environment using 'pip install -r requirements.txt', I got the following error. What can be a possible solution to this dilemma? ``` ERROR:...

Hello, please your help running the Demo: ![image](https://user-images.githubusercontent.com/46006663/228039456-8ce73d6e-804c-43e2-a55d-c0e8546749ce.png) i installed all the requirements. i should change the version of smplx because the new version seems to have an error with...

Hello nkolot, When I try to do some work on humanner.There are one sentence "You can run SMPL-based human pose detectors (e.g., [SPIN], [VIBE], or [ROMP]) on a monocular video...

How to achieve the performance in the paper. I use al l dataset and default options in repo, which cant achieve performance in the paper.