kangya998 / PUGeoNet_pytorch

PUGeoNet_pytorch

Tensorflow version

Data Preparation

download raw mesh data from https://drive.google.com/drive/folders/1n2lf4am9k3hy3ci4W20XiMkXwJKwyg8f?usp=sharing.
run prepare_data/mesh2ply.py and prepare_data/ply2patch.py to prepare the training and testing data.

Training

run main.py to training model with different upsample ratios.

Testing

Upsampling point cloud

run evaluation/eval.py

Evaluation metric

First you need to comple the P2F evaluation.

cd evaluation/p2f  
cmake .  
make  

Then run evaluation/cal_p2f.py and evaluation/cal_metric.py to compute the metric(CD, HD, P2F).