Skip to content

The project is about making a database for an AI about garment recognition. The database is a compact representation of 3D garments worn on the body. This project is the code used to create the database based off of CLOTH3D's database.

Notifications You must be signed in to change notification settings

MLMS-CG/3D-Cloth-Registration

 
 

Repository files navigation

3D-Cloth-Registration

The project is about making a database for an AI about garment recognition. The database is a compact representation of 3D garments worn on the body. This project is the code used to create the database based off of CLOTH3D's database.

Here is the link to my report [UNAVAILABLE], it includes previous reports

Requirements

For the database you might want to download the SMPL bodies from my report and download the CLOTH3D database. For now there is a sample in test_dataset.

To install the requirements, clone the project then write: conda install --file conda_requirements_nvidia.yml conda install --file conda_requirements.yml pip install -r path/to/pip_requirements.txt

Then manually install the remaining libraries by reading the manual requirements file. There will be instructions provided in the file. Those are mostly due to them either not working when installing with conda or pip or because installing manually is the only way.

Run the project

To run the project, the run.py file is given in the code. For now there is a default line inside which will successfully run the project. To test with other files, modify the run.py file to run the project. If you want to turn on testing, uncomment the lines in the "test" section of the file and/or use the object "tests" already given. Just type tests. and different tests will show up. To see the full class, it is main_test,py.

To increase / decrease the logging level, go in main.py line 21 and instead of INFO, write the level you want. Default is INFO.

Testing the project

All tests are in testing/. If you want to add tests to a certain function, open the test_[name] file, with [name] the name of the file that the function is in. For example, testing any functions in nricp.py will be in test_nricp.py. If you add a function make sure to add it within the main_test.py class for it to be run. Then in run.py write tests.test_[module name].py. For example, the module of chamfer.py is distance.

Adding files to the project

If you add a file in the project, make sure to add it in the corresponding module. Then in the module's __init__.py add the following line: from .[new file name] import *. This will ensure that you have access to the file anywhere in the project. To access your file from for example the testing folder, in the test file write from modules.[module name] import [file name]. To access within a neighbour file write from .[file name] import ....

About

The project is about making a database for an AI about garment recognition. The database is a compact representation of 3D garments worn on the body. This project is the code used to create the database based off of CLOTH3D's database.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 100.0%