Replies: 1 comment
|
Same error, were you able to fix it? it seems like 207 (23 * 3 * 3) is the shape of the joint dims but I am not sure what 486 is - 486 comes from the smpl pkl file under the pose dir parameter, and shape dir parameter, but there is no place in the documentastion to know what these things mean, it would be very helpful if you can let me know if you found a solution (since this seems like a 2 year old issue), thank you so much for your time! |
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
Hi,
I installed mmhuman3d, and try to use it but a stack appear when i launch the demo: python demo/estimate_smpl.py configs/hmr/resnet50_hmr_pw3d.py data/checkpoints/resnet50_hmr_pw3d.pth --single_person_demo --det_config demo/mmdetection_cfg/faster_rcnn_r50_fpn_coco.py --det_checkpoint https://download.openmmlab.com/mmdetection/v2.0/faster_rcnn/faster_rcnn_r50_fpn_1x_coco/faster_rcnn_r50_fpn_1x_coco_20200130-047c8118.pth --input_path demo/resources/single_person_demo.mp4 --show_path vis_results/single_person_demo.mp4 --output demo_result --smooth_type savgol --speed_up_type deciwatch --draw_bbox
first:
The model and loaded state dict do not match exactly
second:
File "/workspace/mmhuman3d/demo/estimate_smpl.py", line 591, in
main(args)
File "/workspace/mmhuman3d/demo/estimate_smpl.py", line 489, in main
single_person_with_mmdet(args, frames_iter)
File "/workspace/mmhuman3d/demo/estimate_smpl.py", line 178, in single_person_with_mmdet
mesh_results = inference_image_based_model(
File "/workspace/.miniconda3/lib/python3.10/site-packages/mmhuman3d/apis/inference.py", line 187, in inference_image_based_model
results = model(
File "/workspace/.miniconda3/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1501, in _call_impl
return forward_call(*args, **kwargs)
File "/workspace/.miniconda3/lib/python3.10/site-packages/mmhuman3d/models/architectures/base_architecture.py", line 107, in forward
return self.forward_test(**kwargs)
File "/workspace/.miniconda3/lib/python3.10/site-packages/mmhuman3d/models/architectures/mesh_estimator.py", line 768, in forward_test
pred_output = self.body_model_test(
File "/workspace/.miniconda3/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1501, in _call_impl
return forward_call(*args, **kwargs)
File "/workspace/.miniconda3/lib/python3.10/site-packages/mmhuman3d/models/body_models/smpl.py", line 103, in forward
smpl_output = super(SMPL, self).forward(*args, **kwargs)
File "/workspace/.miniconda3/lib/python3.10/site-packages/smplx/body_models.py", line 369, in forward
vertices, joints = lbs(betas, full_pose, self.v_template,
File "/workspace/.miniconda3/lib/python3.10/site-packages/smplx/lbs.py", line 226, in lbs
pose_offsets = torch.matmul(pose_feature.view(batch_size, -1),
RuntimeError: mat1 and mat2 shapes cannot be multiplied (1x207 and 486x31425)
Do you know how to fix this error ?
All reactions