CAD Object Classification

A Deep-Learning model that can help us to classify 3D object using PoinNet Architecture

Dataset

The dataset was taken from Princeton Modelnet.

Modelnet10 Download

Language & OS

1. Python 3.10.4
2. Windows 10 home

Modules

1. Numpy
2. Matplotlib
3. Tensorflow
4. Pickle
5. Glob
6. trimesh

Installation of modules

Please create a virtual environment then install the modules To create virtual environment open command tab in that folder and run the command

py -3.X venv -m name_of_env

X is refering a python version To activate the virtual environment

.\name_of_env\Scripts\activate

To install all these modules copy the requirement.txt file and run the below command

pip install -r requirment.txt

CAD to Point-Cloud

To Predict 3D object classification we need to transform the 3D Object to a Point-Cloud

Then we will be able to run PointNet Architecture.

Accuracy & Loss

Epoch Size: 50
Batch size: 32
Learning rate: 0.001
Accuracy: 94.03%

Contributing

Pull requests are welcome. For major changes, please open an issue first to discuss what you would like to change.

Please make sure to update tests as appropriate.

License

MIT