Install h5py for Python:
sudo apt-get install libhdf5-dev
sudo pip install h5py
-- Install Chamfer distance package:
cd models/nndistance
python build.py install
Download the MSRA Hand Gesture database at: https://www.dropbox.com/s/bmx2w0zbnyghtp7/cvpr15_MSRAHandGestureDB.zip?dl=0
Unzip it and put folders "P0"~"P9" in the "data/cvpr15_MSRAHandGestureDB" directory.
Please cite "Xiao Sun, Yichen Wei, Shuang Liang, Xiaoou Tang and Jian Sun. Cascaded Hand Pose Regression. IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2015", if you use this database.
cd preprocess
matlab preprocess.m
Please cite this paper if you use this code. "Liuhao Ge, Yujun Cai, Junwu Weng and Junsong Yuan. Hand PointNet: 3D Hand Pose Estimation using Point Sets. In Proc. IEEE Conf. on Computer Vision and Pattern Recognition (CVPR), 2018."
Pretrained Model trained on "P1"~"P9": BaiduNetDesk Link code:z8z8
put encoder model in 'AutoEncoder/results/P0' and regression model in 'results/P0'.
cd AutoEncoder
python train.py
python eval.py
set "capsule model" as the saved model.
python train.py
python eval.py
Our capsule model is based on 3D-point-capsule-networks..