kasvii / PMCE

[ICCV 2023] PyTorch Implementation of "Co-Evolution of Pose and Mesh for 3D Human Body Estimation from Video"
https://kasvii.github.io/PMCE
MIT License
145 stars 5 forks source link

where can I find the 'smpl_mean_vertices.npy'? #3

Closed HenrySomeCode closed 10 months ago

HenrySomeCode commented 10 months ago

When I ran the inference command on colab (I rebuilt this repo on colab):

!python ./main/run_demo.py --vid_file demo/sample_video.mp4 --gpu 0

I get this error:

Traceback (most recent call last):
  File "/content/PMCE/./main/run_demo.py", line 495, in <module>
    main(args)
  File "/content/PMCE/./main/run_demo.py", line 239, in main
    model, joint_regressor, joint_num, skeleton, ckt_name = get_joint_setting(mesh_model, joint_category=joint_set)
  File "/content/PMCE/./main/run_demo.py", line 110, in get_joint_setting
    model = models.PMCE.get_model(num_joint=joint_num, embed_dim=256, depth=3) 
  File "/content/PMCE/main/../lib/models/PMCE.py", line 24, in get_model
    model = PMCE(num_joint, embed_dim, depth)
  File "/content/PMCE/main/../lib/models/PMCE.py", line 13, in __init__
    self.pose_mesh_coevo = CoevoDecoder.get_model(num_joint, embed_dim)
  File "/content/PMCE/main/../lib/models/CoevoDecoder.py", line 250, in get_model
    model = Pose2Mesh(num_joint, embed_dim)
  File "/content/PMCE/main/../lib/models/CoevoDecoder.py", line 200, in __init__
    init_vertices = torch.from_numpy(np.load(SMPL_MEAN_vertices)).cuda()
  File "/usr/local/lib/python3.10/dist-packages/numpy/lib/npyio.py", line 405, in load
    fid = stack.enter_context(open(os_fspath(file), "rb"))
FileNotFoundError: [Errno 2] No such file or directory: 'data/base_data/smpl_mean_vertices.npy'

So where can I find the 'smpl_mean_vertices.npy' ?

kasvii commented 10 months ago

Thanks for pointing out this. Download base_data from OneDrive. Put it below ./data folder.

HenrySomeCode commented 10 months ago

Thanks for pointing out this. Download base_data from OneDrive. Put it below ./data folder.

Thanks! I really appreciate that. I met this error when re-ran the inference command again: image_2023_10_07T05_06_44_994Z For that reason, I changed 'ModelOutput' to 'SMPLX' in smpl_mps.py follow this and I met this error: image

kasvii commented 10 months ago

spin_model_checkpoint.pth.tar has been uploaded to OneDrive. Please download it and put it below ./data/base_data folder.

HenrySomeCode commented 10 months ago

spin_model_checkpoint.pth.tar has been uploaded to OneDrive. Please download it and put it below ./data/base_data folder.

Thanks! It's worked. After adding the spin_model_checkpoint.pth.tar to ./data/base_data I have met this error 😂which said cannot import name 'OSMesaCreateContextAttribs' from 'OpenGL.osmesa'. I uninstalled my pyopengl with version 3.1.0 and re-installed it with version 3.1.5 and finally I got the output!