Errors on Linux (Ubuntu) when running quickstart example

[FrankSpalteholz]

First of all thank you very much for this great repo! When trying the quick-start example i'm facing some errors where i'd really appreciate some help:

1. After processing the videos:

Traceback (most recent call last):
File "scripts/preprocess/extract_video.py", line 267, in
join(args.path, 'openpose_render', sub), args)
File "scripts/preprocess/extract_video.py", line 56, in extract_2d
os.chdir(openpose)
FileNotFoundError: [Errno 2] No such file or directory: '/media/qing/Project/openpose'

2. After triangulation:

-> [Optimize global RT ]: 0.9ms
Traceback (most recent call last):
File "apps/demo/mv1p.py", line 109, in
mv1pmf_smpl(dataset, args)
File "apps/demo/mv1p.py", line 71, in mv1pmf_smpl
weight_shape=weight_shape, weight_pose=weight_pose)
File "/home/frankfurt/dev/EasyMocap/easymocap/pipeline/basic.py", line 77, in smpl_from_keypoints3d2d
params = multi_stage_optimize(body_model, params, kp3ds, kp2ds, bboxes, Pall, weight_pose, cfg)
File "/home/frankfurt/dev/EasyMocap/easymocap/pipeline/basic.py", line 18, in multi_stage_optimize
params = optimizePose3D(body_model, params, kp3ds, weight=weight, cfg=cfg)
File "/home/frankfurt/dev/EasyMocap/easymocap/pyfitting/optimize_simple.py", line 301, in optimizePose3D
params = _optimizeSMPL(body_model, params, prepare_funcs, postprocess_funcs, loss_funcs, weight_loss=weight, cfg=cfg)
File "/home/frankfurt/dev/EasyMocap/easymocap/pyfitting/optimize_simple.py", line 246, in _optimizeSMPL
final_loss = fitting.run_fitting(optimizer, closure, opt_params)
File "/home/frankfurt/dev/EasyMocap/easymocap/pyfitting/optimize.py", line 38, in run_fitting
loss = optimizer.step(closure)
File "/home/frankfurt/dev/EasyMocap/easymocap/pyfitting/lbfgs.py", line 307, in step
orig_loss = closure()
File "/home/frankfurt/dev/EasyMocap/easymocap/pyfitting/optimize_simple.py", line 227, in closure
new_params = func(new_params)
File "/home/frankfurt/dev/EasyMocap/easymocap/pyfitting/optimize_simple.py", line 121, in interp_func
params[key][nf] = interp(params[key][left], params[key][right], 1-weight, key=key)
IndexError: index 800 is out of bounds for dimension 0 with size 800

Then triangulation runs a second time:
(with same error like above) and this one:

Traceback (most recent call last):
File "apps/demo/mv1p.py", line 108, in
mv1pmf_skel(dataset, check_repro=True, args=args)
File "apps/demo/mv1p.py", line 35, in mv1pmf_skel
images, annots = dataset[nf]
File "/home/frankfurt/dev/EasyMocap/easymocap/dataset/mv1pmf.py", line 73, in getitem
annots = self.select_person(annots_all, index, self.pid)
File "/home/frankfurt/dev/EasyMocap/easymocap/dataset/base.py", line 461, in select_person
keypoints = np.zeros((self.config['nJoints'], 3))
KeyError: 'nJoints'

Thank you very much!

[chingswy]

1. After processing the videos:

Did you run this script on zju-ls-feng.zip?
If not, you should install and specify the path by --openpose <openpose_path>.
If so, it doesn't need run OpenPose. Can you show me the folder structure?

[FrankSpalteholz]

Hi Quing, and thank you for the fast reply. Although you were right (haven't copied the openpose-folder to my testdata-folder) it gives me the same error. here is my folder structure:

I'm refering to "mocap_output" as out-folder and "testData" as data-folder.
There is one more thing ... you are mentioning to install 3 smpl-files which are named like "SMPL_FEMALE.pkl" (also for NEUTRAL and MALE) but i've checked all possible downloads on the website "http://smpl.is.tue.mpg.de" but none of them were containing those files. The ones I've found though were inside of "SMPL_python_v.1.1.0.zip" under "models" but their names were different:
basicmodel_f_lbs_10_207_0_v1.1.0.pkl, basicmodel_m_lbs_10_207_0_v1.1.0.pkl and basicmodel_neutral_lbs_10_207_0_v1.1.0.pkl. So although you haven't mentioned so i assumed to just rename them. Maybe that's also causing issues?
Thank you again!

[chingswy]

Download the data with annotation here and extract it. Then run:

data=<path/to/zju-ls-feng-annot>
python3 scripts/preprocess/extract_video.py ${data} --no2d
python3 apps/demo/mv1p.py ${data} --out ${data}/output/smpl --undis --vis_smpl

[FrankSpalteholz]

Thanks again. After following your advice and run the script i got this one:

cameras: 1 10 11 12 13 14 15 16 17 18 19 2 20 21 22 23 3 4 5 6 7 8 9

Demo code for multiple views and one person:

- Input : /home/frankfurt/dev/EasyMocap/data/testData => 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23
- Output: /home/frankfurt/dev/EasyMocap/data/testData/output/smpl
- Body  : smpl=>neutral, body25

triangulation: 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 800/800 [00:04<00:00, 175.61it/s]
dump: 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 800/800 [00:00<00:00, 8895.38it/s]
loading: 100%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 800/800 [00:03<00:00, 212.35it/s]
Traceback (most recent call last):
File "apps/demo/mv1p.py", line 109, in
mv1pmf_smpl(dataset, args)
File "apps/demo/mv1p.py", line 68, in mv1pmf_smpl
body_model = load_model(gender=args.gender, model_type=args.model)
File "/home/frankfurt/dev/EasyMocap/easymocap/smplmodel/body_param.py", line 50, in load_model
regressor_path=reg_path)
File "/home/frankfurt/dev/EasyMocap/easymocap/smplmodel/body_model.py", line 64, in init
self.faces = data['f']
KeyError: 'f'

Edit: i also downloaded the smpl-models for python 1.0.0. But again: they are having different names so i renamed them following your naming-convention. That was correct ... right?

[chingswy]

Can you print data.keys() before this line?

"easymocap/smplmodel/body_model.py", line 64,

[FrankSpalteholz]

loading: 100%|███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 800/800 [00:15<00:00, 52.62it/s]
-> dict_keys(['cam_t', 'poses', 'betas', 'focal_length'])
Traceback (most recent call last):
File "apps/demo/mv1p.py", line 109, in
mv1pmf_smpl(dataset, args)
File "apps/demo/mv1p.py", line 68, in mv1pmf_smpl
body_model = load_model(gender=args.gender, model_type=args.model)
File "/home/frankfurt/dev/EasyMocap/easymocap/smplmodel/body_param.py", line 50, in load_model
regressor_path=reg_path)
File "/home/frankfurt/dev/EasyMocap/easymocap/smplmodel/body_model.py", line 65, in init
self.faces = data['f']
KeyError: 'f'

indicated print(data.keys()) with "->" above

[chingswy]

It's obvious that you used a wrong SMPL model. Please check it.

[FrankSpalteholz]

Thanks but i already wrote that you are using SMPL_FEMALE.pkl (also for NEUTRAL and MALE) but on the website you've mentioned there are no such models available with these names. I was downloading SMPL_python_v.1.0.0.zip (containing basicmodel_m_lbs_10_207_0_v1.0.0.pkl and basicModel_f_lbs_10_207_0_v1.0.0.pkl) and mpips-smplify_public_v2.zip (containing basicModel_neutral_lbs_10_207_0_v1.0.0.pkl) These are the ones i'm using (i've renamed them to SMPL_FEMALE.pkl etc to follow your naming-convention) but if these are not the correct ones where can i find them?
Thank you so much again for your help

[chingswy]

To check your model, run this:

import pickle
with open('basicmodel_m_lbs_10_207_0_v1.0.0.pkl', 'rb') as f:
    model = pickle.load(f, encoding='latin1') 
    print(model.keys())

The output of mine is:

dict_keys(['J_regressor_prior', 'f', 'J_regressor', 'kintree_table', 'J', 'weights_prior', 'weights', 'vert_sym_idxs', 'posedirs', 'pose_training_info', 'bs_style', 'v_template', 'shapedirs', 'bs_type'])

You can set other gender when running the demo code:

python3 apps/demo/mv1p.py ${data} --out ${data}/output/smpl --undis --vis_smpl --gender male
python3 apps/demo/mv1p.py ${data} --out ${data}/output/smpl --undis --vis_smpl --gender female

[FrankSpalteholz]

Excellent ... my bad (of course) Thanks a lot for having so much patience with me!