I try to run the demo on Google collab, but the following errors always appear. All processes are carried out according to the video, and the input image is also used test.png Why do the following errors occur? After trying many solutions, I don't know if anyone has the same problem. Can it be solved?

TypeError Traceback (most recent call last) in () 3 load_state(net, checkpoint) 4 ----> 5 get_rect(net.cuda(), [image_path], 512)

in get_rect(net, images, height_size) 28 total_keypoints_num += extract_keypoints(heatmaps[:, :, kpt_idx], all_keypoints_by_type, total_keypoints_num) 29 ---> 30 pose_entries, all_keypoints = group_keypoints(all_keypoints_by_type, pafs, demo=True) 31 for kpt_id in range(all_keypoints.shape[0]): 32 all_keypoints[kpt_id, 0] = (all_keypoints[kpt_id, 0] * stride / upsample_ratio - pad[1]) / scale

TypeError: group_keypoints() got an unexpected keyword argument 'demo'

It was running fine, just 2-3 hours ago, but isn't working now. Probably there has been some updates on the github repositories used in this notebook.

Complete error log:

---------------------------------------------------------------------------
ImportError                               Traceback (most recent call last)
<ipython-input-14-22614b7ef849> in <module>()
----> 1 from lib.colab_util import generate_video_from_obj, set_renderer, video
      2 
      3 renderer = set_renderer()
      4 generate_video_from_obj(obj_path, out_img_path, video_path, renderer)
      5 

4 frames
/content/pifuhd/lib/colab_util.py in <module>()
     33 
     34 # Util function for loading meshes
---> 35 from pytorch3d.io import load_objs_as_meshes
     36 
     37 from IPython.display import HTML

/usr/local/lib/python3.6/dist-packages/pytorch3d/io/__init__.py in <module>()
      2 
      3 
----> 4 from .obj_io import load_obj, load_objs_as_meshes, save_obj
      5 from .ply_io import load_ply, save_ply
      6 

/usr/local/lib/python3.6/dist-packages/pytorch3d/io/obj_io.py in <module>()
     12 from pytorch3d.io.mtl_io import load_mtl, make_mesh_texture_atlas
     13 from pytorch3d.io.utils import _check_faces_indices, _make_tensor, _open_file
---> 14 from pytorch3d.renderer import TexturesAtlas, TexturesUV
     15 from pytorch3d.structures import Meshes, join_meshes_as_batch
     16

/usr/local/lib/python3.6/dist-packages/pytorch3d/renderer/__init__.py in <module>()
      1 # Copyright (c) Facebook, Inc. and its affiliates. All rights reserved.
      2 
----> 3 from .blending import (
      4     BlendParams,
      5     hard_rgb_blend,

/usr/local/lib/python3.6/dist-packages/pytorch3d/renderer/blending.py in <module>()
      7 
      8 # pyre-fixme[21]: Could not find name `_C` in `pytorch3d`.
----> 9 from pytorch3d import _C
     10 
     11 

ImportError: /usr/local/lib/python3.6/dist-packages/pytorch3d/_C.cpython-36m-x86_64-linux-gnu.so: undefined symbol: _ZN3c104impl23ExcludeDispatchKeyGuardC1ENS_11DispatchKeyE

---------------------------------------------------------------------------
NOTE: If your import is failing due to a missing package, you can
manually install dependencies using either !pip or !apt.

To view examples of installing some common dependencies, click the
"Open Examples" button below.
--------------------------------------------------------------------------- 

ImportError Traceback (most recent call last) in () ----> 1 from lib.colab_util import generate_video_from_obj, set_renderer, video 2 3 renderer = set_renderer() 4 generate_video_from_obj(obj_path, out_img_path, video_path, renderer) 5

/content/pifuhd/lib/colab_util.py in () 39 40 # Data structures and functions for rendering ---> 41 from pytorch3d.structures import Meshes, Textures 42 from pytorch3d.renderer import ( 43 look_at_view_transform,

ImportError: cannot import name 'Textures'

from the google colab could be due to pytorch3d version?

Which file from the openpose binaries should I use on Windows10? The file (./build/examples/openpose/openpose.bin) given in batch_openpose.py does not exist on the Windows 10 binaries with the openpose version 1.6.0

Btw:. In the documentation the command (python apps/process_openpose.py .....) is wrong. It has to be (python apps/batch_openpose.py .....) i guess.

Thank you for your great work

but i have a trouble with getting color mesh

I changed from get_mesh to get_mesh_img on recon.py and executed. The result is the same on the front and back as in the picture below. Is there any additional part that I need to set up?

Thank you

When I read the code of HGPIFuMRNet.py, I'm not clear about the meaning of images_local([B1, B2, C, H, W]) and images_global([B1, C, H, W]). What is the difference between images_local and images_global and what does B1, B2 mean? And for points_nml and labels_nml, are they from .obj file or calculated? @shunsukesaito Thank you!

Google Colab notebook does not work with an error. When I try to move the second Render the result block, I get the following error:

ImportError: /usr/local/lib/python3.6/dist-packages/pytorch3d/_C.cpython-36m-x86_64-linux-gnu.so: undefined symbol: _ZN3c104impl23ExcludeDispatchKeyGuardC1ENS_11DispatchKeyE

env: pytorch1.7.0、 cuda 11.0、RTX2060 、win10system.

get this Error: RuntimeError: CUDA out of memory. Tried to allocate 256.00 MiB (GPU 0; 6.00 GiB total capacity; 3.87 GiB already allocated; 187.62 MiB free; 4.01 GiB reserved in total by PyTorch)

is there any method to solve this problem?

i have a question about HGPIFuNetwNML.py code，why the def get_error has a parameter named gamma，the have bug when i run the code，because the loss only need pred and label that two input，and the loss function is defined to MSE ，i see the paper about the loss function is BCE

The number of network parameters of PIFuHD is 387 million,which is 25 times that of PIFu. Can you share with us the hardware requirements of your GPU when training pifuhd? Such as GPU memory, type and number of gpus.The GPU memory I used for training was 24g,batch_size=1, but still cuda out of memory

Hi @shunsukesaito

I could run the code on colab on my test data and worked fine. I am trying to run it on gpu machine getting this following error.

Traceback (most recent call last): File "/home/hamid_farhid/anaconda3/envs/PIFuHD/lib/python3.7/runpy.py", line 193, in _run_module_as_main "main", mod_spec) File "/home/hamid_farhid/anaconda3/envs/PIFuHD/lib/python3.7/runpy.py", line 85, in _run_code exec(code, run_globals) File "/home/hamid_farhid/PIFuHD/pifuhd/apps/simple_test.py", line 30, in reconWrapper(cmd, args.use_rect) File "/home/hamid_farhid/PIFuHD/pifuhd/apps/recon.py", line 220, in reconWrapper recon(opt, use_rect) File "/home/hamid_farhid/PIFuHD/pifuhd/apps/recon.py", line 210, in recon gen_mesh(opt.resolution, netMR, cuda, test_data, save_path, components=opt.use_compose) File "/home/hamid_farhid/PIFuHD/pifuhd/apps/recon.py", line 38, in gen_mesh net.filter_global(image_tensor_global) File "/home/hamid_farhid/PIFuHD/pifuhd/lib/model/HGPIFuMRNet.py", line 83, in filter_global self.netG.filter(images) File "/home/hamid_farhid/PIFuHD/pifuhd/lib/model/HGPIFuNetwNML.py", line 134, in filter self.im_feat_list, self.normx = self.image_filter(images) File "/home/hamid_farhid/anaconda3/envs/PIFuHD/lib/python3.7/site-packages/torch/nn/modules/module.py", line 489, in call result = self.forward(*input, **kwargs) File "/home/hamid_farhid/PIFuHD/pifuhd/lib/model/HGFilters.py", line 195, in forward hg = self._modules'm' + str(i) File "/home/hamid_farhid/anaconda3/envs/PIFuHD/lib/python3.7/site-packages/torch/nn/modules/module.py", line 489, in call result = self.forward(*input, **kwargs) File "/home/hamid_farhid/PIFuHD/pifuhd/lib/model/HGFilters.py", line 117, in forward return self._forward(self.depth, x) File "/home/hamid_farhid/PIFuHD/pifuhd/lib/model/HGFilters.py", line 103, in _forward low2 = self._forward(level - 1, low1) File "/home/hamid_farhid/PIFuHD/pifuhd/lib/model/HGFilters.py", line 111, in _forward up2 = F.interpolate(low3, scale_factor=2, mode='bicubic', align_corners=True) File "/home/hamid_farhid/anaconda3/envs/PIFuHD/lib/python3.7/site-packages/torch/nn/functional.py", line 2459, in interpolate " (got {})".format(input.dim(), mode)) NotImplementedError: Input Error: Only 3D, 4D and 5D input Tensors supported (got 4D) for the modes: nearest | linear | bilinear | trilinear (got bicubic) freeglut (foo): failed to open display ''

By referencing this: Error, we should do some changes on input tensor. Any thought on this one?

I have followed all the steps from this video: https://www.youtube.com/watch?v=QANxO69y6r4 But after having the pifuhd file appear and the lightweight-human-pose estimation file appear, the results file was empty and just had the recon file. I had no errors when running, but this somehow happened. Any help would be appreciated.

I installed all the libraries successfully but I got this unexpected kill process without generating the 3d mesh

Error: generate mesh (test) ... 0%| | 0/1 [00:00<?, ?it/s]./results/pifuhd_final/recon/result_test_512.obj Killed []

I tried to make the code work all the ways I could and it simply did not load I got the stat dict error several times, I think it doesn't work anymore

Hi All,

Hope you are doing well.

I am newbie, so please bear with me. Also, let me know what further information is required so that you can help me with solutions.

Please refer to the attached png, can anyone tell me why the render is broken?

Many thanks in advance.

Regards