Jeff-sjtu / HybrIK

Official code of "HybrIK: A Hybrid Analytical-Neural Inverse Kinematics Solution for 3D Human Pose and Shape Estimation", CVPR 2021
MIT License
1.17k stars 142 forks source link

'<=' not supported between instances of 'tuple' and 'int' #104

Open hzlhzlhzlhzl opened 1 year ago

hzlhzlhzlhzl commented 1 year ago

0% 0/10 [00:00<?, ?it/s] Traceback (most recent call last): File "scripts/demo_image.py", line 144, in focal_length=focal, height=image.shape[0], width=image.shape[1]) File "/content/HybrIK/hybrik/utils/render_pytorch3d.py", line 81, in render_mesh imgs = renderer(mesh) File "/usr/local/envs/hybrik/lib/python3.7/site-packages/torch/nn/modules/module.py", line 722, in _call_impl result = self.forward(input, kwargs) File "/usr/local/envs/hybrik/lib/python3.7/site-packages/pytorch3d/renderer/mesh/renderer.py", line 54, in forward fragments = self.rasterizer(meshes_world, kwargs) File "/usr/local/envs/hybrik/lib/python3.7/site-packages/torch/nn/modules/module.py", line 722, in _call_impl result = self.forward(input, **kwargs) File "/usr/local/envs/hybrik/lib/python3.7/site-packages/pytorch3d/renderer/mesh/rasterizer.py", line 151, in forward cull_backfaces=raster_settings.cull_backfaces, File "/usr/local/envs/hybrik/lib/python3.7/site-packages/pytorch3d/renderer/mesh/rasterize_meshes.py", line 115, in rasterize_meshes if image_size <= 64: TypeError: '<=' not supported between instances of 'tuple' and 'int' 我在运行colab时出现的,一开始是在scipy==1.1.0时出现No module named 'scipy.spatial.transform',我把scipy更新后出现上述bug

timbsb commented 1 year ago

I was also confused yesterday. now I solved this problem. It is because the version of PyTorch is too old. pip can install version 0.3.0 only. I use Conda to solve this problem like this:

!source activate hybrik && conda install -c pytorch3d pytorch3d -y