The code of 3D Identification
Linux16.04/Windows10/MacOS10.15.5 mex Compiler : Clang MatLab : R2020b
We will introduce several directories.
- core: the core function about alignment and split strategies.
- example: some examples of how to use our method.
- gaussian: the code of resisting gaussian noise.
- mex: some C++ dependencies which need to be compiled before using our method.
- model: some models provided for tests.
- retrieval: the code used to retrieve 4 dataset.
- segmentation: the code showing the effectiveness of segmentation strategies.
- thirdparty: the thirdparty code used in our method.
- util: some utilization function used in our method such as reading mesh.
- watermarking: the code used to test watermarking models.
- setup.m: used to prepare environment: compile dependencies.
- experiments: the retrieve intermediate data generated during the experiment.
- tools: tools to compute top rate.
To run the code in examples, you should call setup.m
firstly.
Then just use the .m file in fold example.
For example, input dragon_rh
command in matlab window.
Each sub-directory in /retrieval has two files. One is used to prepare data (generate cvs), and the other is used to retrieve dataset.
We did some extra experiments on PointNetVLAD Datasets. And the code is in the file /util/retrieve_residual.m. You should download the benchmark dataset first when try to do it. And you should modify the dataset path in /util/retrieve_residual.m.