3d-gan

Training:

1. Prepare voxel data and keep them in similar structure as data folder. The 'data' folder is a sample of data directory structure. Here I have used ShapeNet dataset with chair object.
2. If you have .obj data you can convert them into .binvox by gen_voxels.py
3. Set proper parameters in config.py.
4. To train use train_3D-GAN.py

Inference

provide the path of the generator model and run inference.py

Sample output