Output_Folder only contains a empty folder

[lunacy-xc]

And when I set --show to be true , some errors occur . OpenGL.error.GLError: GLError(err = 12289, baseOperation = eglInitialize, cArguments = ( <OpenGL._opaque.EGLDisplay_pointer object at 0x7f198c31fa70>, c_long(0), c_long(0), ), result = 0 )

And I think it related to openGL . I have s 3.3 version opengl and can you tell me your version ?

[vchoutas]

What is your environment? Have you tried setting your EGL device ID with the env var EGL_DEVICE_ID ?

[shriyakak]

@vchoutas: OUTPUT_FOLDER contains "man-in-red-crew-neck-sweatshirt-photography-941693.png_000" folder not an image. I'm not getting any errors while installing prereqs excepts some warnings as below. could you please confirm how should be the output generated? I tried running it on google colab and Ubuntu 18.04, Cuda 10.1, torch 1.5.

"_The following keys were not found: ['smplx.head_idxs', 'smplx.body_model.left_hand_components', 'smplx.body_model.right_hand_components', 'smplx.body_model.left_hand_mean', 'smplx.body_model.right_hand_mean', 'smplx.body_model.pose_mean', 'smplx.body_model.dynamic_lmk_bary_coords'] 2020-10-12 08:20:52.434 | WARNING | expose.utils.checkpointer:load_checkpoint:125 - The following keys were not expected: ['smplx.body_model.source_idxs', 'smplx.body_model.target_idxs', 'smplx.body_model.extra_joint_regressor', 'smplx.body_model.dynamic_lmk_b_coords', 'smplx.hand_predictor.hand_offset', 'smplx.hand_predictor.hand_model.extra_joints_idxs', 'smplx.hand_predictor.hand_model.faces_tensor', 'smplx.hand_predictor.hand_model.v_template', 'smplx.hand_predictor.hand_model.shapedirs', 'smplx.hand_predictor.hand_model.J_regressor', 'smplx.hand_predictor.hand_model.posedirs', 'smplx.hand_predictor.hand_model.parents', 'smplx.hand_predictor.hand_model.lbs_weights', 'smplx.hand_predictor.pca_decoder.pca_basis', 'smplx.hand_predictor.pca_decoder.inv_pca_basis', 'smplx.hand_predictor.pca_decoder.mean', 'smplx.head_predictor.head_offset', 'smplx.head_predictor.head_vertices_ids', 'smplx.head_predictor.head_model.faces_tensor', 'smplx.head_predictor.head_model.v_template', 'smplx.head_predictor.head_model.shapedirs', 'smplx.head_predictor.head_model.expr_dirs', 'smplx.head_predictor.head_model.J_regressor', 'smplx.head_predictor.head_model.posedirs', 'smplx.head_predictor.head_model.parents', 'smplx.head_predictor.head_model.lbs_weights', 'smplx.head_predictor.head_model.lmk_faces_idx', 'smplx.head_predictor.head_model.lmk_bary_coords', 'smplx.head_predictor.head_model.dynamic_lmk_faces_idx', 'smplx.head_predictor.head_model.dynamic_lmk_b_coords', 'smplx.head_predictor.head_model.neck_kin_chain', 'smplx.body_loss.edge_loss.gt_connections', 'smplx.body_loss.edge_loss.est_connections', 'smplx.hand_loss.edge_loss.gt_connections', 'smplx.hand_loss.edge_loss.est_connections', 'smplx.head_loss.edge_loss.gt_connections', 'smplx.head_loss.edge_loss.estconnections'] 0% 0/1 [00:00<?, ?it/s]"

[vchoutas]

The warnings can be safely ignored. Was anything else printed? What code did you use in Colab?

[bhumikasinghrk]

Same Issue, when --show = False, this error occurs:

File "demo.py", line 565, in rcnn_batch=rcnn_batch, File "/usr/local/lib/python3.6/dist-packages/torch/autograd/grad_mode.py", line 26, in decorate_context return func(*args, **kwargs) File "demo.py", line 366, in main bg_imgs=bg_hd_imgs, UnboundLocalError: local variable 'bg_hd_imgs' referenced before assignment

and when --show = True, OUTPUT_FOLDER contains "man-in-red-crew-neck-sweatshirt-photography-941693.png_000" folder not an image.

I tried this both in colab and with AWS, I am getting the same error.

[boyisolation]

Encountered the same problem upstairs show--false will show a empty folder, show--true will show a error :UnboundLocalError: local variable 'bg_hd_imgs' referenced before assignment这样的错误

[boyisolation]

@vchoutas：OUTPUT_FOLDER包含“ man-in-red-crew-neck-sweatshirt-photography-941693.png_000”文件夹，而不是图像。除以下一些警告外，安装prereqs时没有任何错误。您能否确认应该如何生成输出？ 我尝试在Google colab和Ubuntu 18.04，Cuda 10.1，Torch 1.5上运行它。

“_未找到以下键：['smplx.head_idxs'，'smplx.body_model.left_hand_components'，'smplx.body_model.right_hand_components'，'smplx.body_model.left_hand_mean'，'smplx.body_model.right_hand_mean'，'smplx。 body_model.pose_mean'，'smplx.body_model.dynamic_lmk_bary_coords'] 2020-10-12 08：20：52.434 | 警告| Exposure.utils.checkpointer：load_checkpoint：125-不需要以下键：['smplx.body_model.source_idxs'，'smplx.body_model.target_idxs'，'smplx.body_model.extra_joint_regressor'，'smplx.body_model.dynamic_lmk_b_coords' 'smplx.hand_predictor.hand_offset'，'smplx.hand_predictor.hand_model.extra_joints_idxs'，'smplx.hand_predictor.hand_model.faces_tensor'，'smplx.hand_predictor.hand_model.v_template'，'smplx.hand_pre_predictor。 .hand_predictor.hand_model.J_regressor'，'smplx.hand_predictor.hand_model.posedirs'，'smplx.hand_predictor.hand_model.parents'，'smplx.hand_predictor.hand_model.lbs_weights'，'smplx.hand_predictor.pcais'。 。 0％0/1 [00:00 <?,？it / s]”

你好我遇到相同的问题，请问你有解决方案吗

[Anirudh257]

@shriyakak you can't run it on colab. Open3D requires a server. You need to run it on a desktop with GPU installed.

[melih-unsal]

I have also encountered on my local environment. When show=True it gives me the same error above. Is there anyone fixed it?

[booker-max]

Do you have the training code for this paper? Can you share it with me? This is my email address 1654617276wj@gmail.com，Thank you very much.！

[Kaius96]

Was this issue resolved? @lunacy-xc Did you fix the empty output folder? If so please let me know how, I have the same issue. Setting the EGL_DEVICE_ID=1 in the Enviroment Variables did not help.

[Anirudh257]

@Kaius96 I wrote a workaround code for this problem here: https://github.com/Anirudh257/Render-SMPL-X-mesh-on-system-without-GPU Let me know if it works for you.

[pyxploiter]

This works for me. saves the output in the folder (don't use --show parameters)

python demo.py --image-folder samples --exp-cfg data/conf.yaml --save-vis true