Closed sarahxinxi closed 3 years ago
Hi @sarahxinxi , What do you mean by wrong?
This demo is a unit test. It samples pose and shape parameters randomly from a gaussian distribution and compare the output from all frameworks. The output mesh will not look sensible, because the input is not sensible.
Can I close this issue now?
Hi @sarahxinxi , What do you mean by wrong?
This demo is a unit test. It samples pose and shape parameters randomly from a gaussian distribution and compare the output from all frameworks. The output mesh will not look sensible, because the input is not sensible.
Can I close this issue now?
So how can I get a sensible output mesh with a pose not in a T-pose
Hi @sarahxinxi , you can fit STAR to SMPL meshes from the amass dataset https://amass.is.tue.mpg.de/
In compare_frameworks.py, I want to see the result of mesh. I write obj result with tf_star,d and model.r respectively, all these results are wrong, but obj result with v_shaped is right. What's wrong