I am getting the following error when I try to run training, how should I proceed in order to solve it?

(mvp) jpsml@jpsml-ubuntu:~/mvp$ python -m torch.distributed.launch --nproc_per_node=8 --use_env run/train_3d.py --cfg configs/campus/mvp_campus.yaml

Setting OMP_NUM_THREADS environment variable for each process to be 1 in default, to avoid your system being overloaded, please further tune the variable for optimal performance in your application as needed.

Traceback (most recent call last): File "run/train_3d.py", line 34, in import dataset File "/home/jpsml/mvp/run/../lib/dataset/init.py", line 20, in from dataset.h36m import H36M as h36m File "/home/jpsml/mvp/run/../lib/dataset/h36m.py", line 30, in from lib.utils.cameras_cpu import camera_to_world_frame, project_pose ModuleNotFoundError: No module named 'lib' Traceback (most recent call last): File "run/train_3d.py", line 34, in import dataset File "/home/jpsml/mvp/run/../lib/dataset/init.py", line 20, in from dataset.h36m import H36M as h36m File "/home/jpsml/mvp/run/../lib/dataset/h36m.py", line 30, in from lib.utils.cameras_cpu import camera_to_world_frame, project_pose ModuleNotFoundError: No module named 'lib' Traceback (most recent call last): File "run/train_3d.py", line 34, in import dataset File "/home/jpsml/mvp/run/../lib/dataset/init.py", line 20, in from dataset.h36m import H36M as h36m File "/home/jpsml/mvp/run/../lib/dataset/h36m.py", line 30, in from lib.utils.cameras_cpu import camera_to_world_frame, project_pose ModuleNotFoundError: No module named 'lib' Traceback (most recent call last): File "run/train_3d.py", line 34, in import dataset File "/home/jpsml/mvp/run/../lib/dataset/init.py", line 20, in from dataset.h36m import H36M as h36m File "/home/jpsml/mvp/run/../lib/dataset/h36m.py", line 30, in from lib.utils.cameras_cpu import camera_to_world_frame, project_pose ModuleNotFoundError: No module named 'lib' Traceback (most recent call last): File "run/train_3d.py", line 34, in import dataset File "/home/jpsml/mvp/run/../lib/dataset/init.py", line 20, in from dataset.h36m import H36M as h36m File "/home/jpsml/mvp/run/../lib/dataset/h36m.py", line 30, in from lib.utils.cameras_cpu import camera_to_world_frame, project_pose ModuleNotFoundError: No module named 'lib' Traceback (most recent call last): File "run/train_3d.py", line 34, in import dataset File "/home/jpsml/mvp/run/../lib/dataset/init.py", line 20, in from dataset.h36m import H36M as h36m File "/home/jpsml/mvp/run/../lib/dataset/h36m.py", line 30, in from lib.utils.cameras_cpu import camera_to_world_frame, project_pose ModuleNotFoundError: No module named 'lib' Traceback (most recent call last): File "run/train_3d.py", line 34, in import dataset File "/home/jpsml/mvp/run/../lib/dataset/init.py", line 20, in from dataset.h36m import H36M as h36m File "/home/jpsml/mvp/run/../lib/dataset/h36m.py", line 30, in from lib.utils.cameras_cpu import camera_to_world_frame, project_pose ModuleNotFoundError: No module named 'lib' Traceback (most recent call last): File "run/train_3d.py", line 34, in import dataset File "/home/jpsml/mvp/run/../lib/dataset/init.py", line 20, in from dataset.h36m import H36M as h36m File "/home/jpsml/mvp/run/../lib/dataset/h36m.py", line 30, in from lib.utils.cameras_cpu import camera_to_world_frame, project_pose ModuleNotFoundError: No module named 'lib' Traceback (most recent call last): File "/home/jpsml/anaconda3/envs/mvp/lib/python3.6/runpy.py", line 193, in _run_module_as_main "main", mod_spec) File "/home/jpsml/anaconda3/envs/mvp/lib/python3.6/runpy.py", line 85, in _run_code exec(code, run_globals) File "/home/jpsml/anaconda3/envs/mvp/lib/python3.6/site-packages/torch/distributed/launch.py", line 261, in main() File "/home/jpsml/anaconda3/envs/mvp/lib/python3.6/site-packages/torch/distributed/launch.py", line 257, in main cmd=cmd) subprocess.CalledProcessError: Command '['/home/jpsml/anaconda3/envs/mvp/bin/python', '-u', 'run/train_3d.py', '--cfg', 'configs/campus/mvp_campus.yaml']' returned non-zero exit status 1.

When I run the validation code with given pretrained checkpoint on Panoptic dataset, I found that there is something wrong with batch size larger than 1. For example, When batch size=1, the precision can be reproduced.

When batch size=2, it seems that the model fails to predict correctly.

Does anyone get the similar problem? Or can you please give some advice about the reason of this problem? thx!

I have a question about your work. "Direct Multi-view Multi-person 3D Pose Estimation NIPS 2021". Your multiview performance on Panoptic Dataset is much better than VoxelPose. However, why aren't you as good as him with a single view setting. Your MPJPE is 93.8mm while VoxelPose's MPJPE is 66.95mm. And of course, I can't reproduce their results. Could you help me with this problem? Thx

Hi, I encountered the following error when training the first epoch in the evaluation. Could you help find out the problem? Thanks in anvance.

INFO:core.function:Test: [200/323]      Time: 0.178s (0.291s)   Speed: 28.1 samples/s   Data: 0.000s (0.055s)   Memory 465635328.0
Traceback (most recent call last):
  File "run/train_3d.py", line 334, in <module>
    main()
  File "run/train_3d.py", line 260, in main
    final_output_dir, thr, num_views=num_views)
  File "/mnt/lustre/liqikai.vendor/open_mmlab/pose3d/mvp/lib/core/function.py", line 161, in validate_3d
    for i, (inputs, meta) in enumerate(loader):
  File "/mnt/lustre/liqikai.vendor/anaconda3/envs/pt180cu111py37mmcv1317/lib/python3.7/site-packages/torch/utils/data/dataloader.py", line 517, in __next__
    data = self._next_data()
  File "/mnt/lustre/liqikai.vendor/anaconda3/envs/pt180cu111py37mmcv1317/lib/python3.7/site-packages/torch/utils/data/dataloader.py", line 1179, in _next_data
    return self._process_data(data)
  File "/mnt/lustre/liqikai.vendor/anaconda3/envs/pt180cu111py37mmcv1317/lib/python3.7/site-packages/torch/utils/data/dataloader.py", line 1225, in _process_data
    data.reraise()
  File "/mnt/lustre/liqikai.vendor/anaconda3/envs/pt180cu111py37mmcv1317/lib/python3.7/site-packages/torch/_utils.py", line 429, in reraise
    raise self.exc_type(msg)
ValueError: Caught ValueError in DataLoader worker process 1.
Original Traceback (most recent call last):
  File "/mnt/lustre/liqikai.vendor/anaconda3/envs/pt180cu111py37mmcv1317/lib/python3.7/site-packages/torch/utils/data/_utils/worker.py", line 202, in _worker_loop
    data = fetcher.fetch(index)
  File "/mnt/lustre/liqikai.vendor/anaconda3/envs/pt180cu111py37mmcv1317/lib/python3.7/site-packages/torch/utils/data/_utils/fetch.py", line 44, in fetch
    data = [self.dataset[idx] for idx in possibly_batched_index]
  File "/mnt/lustre/liqikai.vendor/anaconda3/envs/pt180cu111py37mmcv1317/lib/python3.7/site-packages/torch/utils/data/_utils/fetch.py", line 44, in <listcomp>
    data = [self.dataset[idx] for idx in possibly_batched_index]
  File "/mnt/lustre/liqikai.vendor/open_mmlab/pose3d/mvp/lib/dataset/panoptic.py", line 257, in __getitem__
    i, m = super().__getitem__(self.num_views * idx + k)
ValueError: too many values to unpack (expected 2)

Hello, Can I use your model in my configuration , say 3-4 cameras, by using my cameras' params ? Have you tested the method for its inference speed and if so how fast is it ? Thanks

In best_model_config.yaml num_feature_levels: 1 In config.py config.DECODER.use_feat_level = [0, 1, 2]

I am confused about that if use_feat_level = [0, 1, 2], should the num_feature_levels be equal to 3 ? thanks

Thanks for your excellent. I noticed that you showed the result on Human3.6M dataset, and compared with voxelpose on the same dataset, but related files can not be found. Could you please share the trained checkpoint or training configuration? And how long have you trained MvP and Voxelpose to get final result, respectively? Thanks for your help!

Thank you very much for your perfect work. I thind that the visualizations look cool. I would like to ask how to visualize the results as yours.

Thanks again.

Dear authors,

In lib.models/ops/modules/projattn.py, I noticed that the weights of self.sampling_offsets is set to constant 0, and the bias has no gradient backpropagation (line 94- line 105).

https://github.com/sail-sg/mvp/blob/80eecd012f51f49da357e337716d40a6398d520d/lib/models/ops/modules/projattn.py#L94

In my opinion, if the weights are set to 0 and the bias has no gradient, the sampled offsets will be always the same across different training samples. But it seems the offsetted points are informatively selected according to Figure 5 in your paper.

On the other hand, in the provided pretrained model, the weights and the bias are different from what they are initialized. Could you please tell me what is the final initialization method of self.sampling_offsets? Thank you very much!

When I prepare to run experiments on Human 3.6M dataset, I found that the getitem function calls the function from the class it inheritates, which is JointDataset, but JointDateset return 2 items (images and meta info) only, while human 3.6m requires 5 items. Is there any reference to use the human36m dataset? Thanks

Hi, I was trying to run a quick evaluation with the provided pre-trained model for the campus dataset. However, it seems that the pre-trained weights (d1_384_85.2.pth.tar) do not match the model. Can you help to double-check the provided pre-trained weight file? Thank you so much.

In your paper, you mention that you have replaced the projective attention with dense attention module, here is the results:

I wonder how did you run the experiment? How can I modify your code to run the experiment? Which module should I modify?