A library for using motion capture data for machine learning
This library is currently highly experimental and everything is subject to change :)
- Mocap Data Parsers and Writers
- Common mocap pre-processing algorithms
- Feature extraction library
- Visualization tools
- Read BVH Files
- Write BVH Files
- Pre-processing pipelines
- Supporting
scikit-learnAPI - Convert data representations
- Euler angles to positions
- Euler angles to exponential maps
- Exponential maps to euler angles
- Body-oriented global translation and rotation calculation with inverse tranform
- Root-centric position normalizer with inverse tranform
- Standard scaler
- Joint selectors
- Supporting
- Annotations
- Foot/ground contact detector
from pymo.parsers import BVHParser
parser = BVHParser()
parsed_data = parser.parse('demos/data/AV_8Walk_Meredith_HVHA_Rep1.bvh')from pymo.preprocessing import *
from sklearn.pipeline import Pipeline
data_pipe = Pipeline([
('param', MocapParameterizer('position')),
('rcpn', RootCentricPositionNormalizer()),
('delta', RootTransformer('abdolute_translation_deltas')),
('const', ConstantsRemover()),
('np', Numpyfier()),
('down', DownSampler(2)),
('stdscale', ListStandardScaler())
])
piped_data = data_pipe.fit_transform([parsed_data])mp = MocapParameterizer('position')
positions = mp.fit_transform([parsed_data])from pymo.features import *
plot_foot_up_down(positions[0], 'RightFoot_Yposition')signal = create_foot_contact_signal(positions[0], 'RightFoot_Yposition')
plt.figure(figsize=(12,5))
plt.plot(signal, 'r')
plt.plot(positions[0].values['RightFoot_Yposition'].values, 'g')For any questions, feedback, and bug reports, please use the Github Issues.
Created by Omid Alemi
This code is available under the MIT license.