Are you sure you want to delete this task? Once this task is deleted, it cannot be recovered.
You can not select more than 25 topics
Topics must start with a chinese character,a letter or number, can include dashes ('-') and can be up to 35 characters long.
zhangych02
25027f3729
|
1 year ago | |
|---|---|---|
| src | 2 years ago | |
| torch | 2 years ago | |
| LICENSE | 1 year ago | |
| PUGAN.pdf | 2 years ago | |
| README.md | 1 year ago | |
| logo.png | 1 year ago | |
| performance.jpg | 1 year ago | |
| pugan algorithm analysis.docx | 2 years ago | |
README.md
PU-GAN: a Point Cloud Upsampling Adversarial Network.
PU-GAN is a point cloud upsampling algorithm collected into our open source algorithm library of point cloud. The official code is tested under TF1.11 and Python 3.6 on Ubuntu 16.04, which is stored in folder './src'. In this project, we expande the PyTorch version of PU-GAN on python 3.6 and PyTorch 1.12, which is in folder './torch'.
Algorithm analysis
please refer to "pugan algorithm analysis.doc"
Running project
please refer to README.md in each version.
Performance comparison
We compare the algorithm performance of different frameworks under tensorflow and pytorch respectively. Because the test data is not provided in source project, we choose two point cloud objects from 3PU and only calculate pdf values duing to the lack of ground truth.
Contributors
name: Zhang Yongchi
email: zhangych02@pcl.ac.cn
No Description
Text Makefile C++ Python C other
MIT