This repository is an attempt to develop the object classifier part of the PointNet research paper. The classifier is implemented using TensorFlow.
The classifier classifies between 40 objects, which are,
- airplane
- bathtub
- bed
- bench
- bookshelf
- bottle
- bowl
- car
- chair
- cone
- cup
- curtain
- desk
- door
- dresser
- flower_pot
- glass_box
- guitar
- keyboard
- lamp
- laptop
- mantel
- monitor
- night_stand
- person
- piano
- plant
- radio
- range_hood
- sink
- sofa
- stairs
- stool
- table
- tent
- toilet
- tv_stand
- vase
- wardrobe
- xbox
@article{qi2016pointnet,
title={PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation},
author={Qi, Charles R and Su, Hao and Mo, Kaichun and Guibas, Leonidas J},
journal={arXiv preprint arXiv:1612.00593},
year={2016}
}
- Python3
- TensofFlow
- Numpy
- MobileNet40 dataset
To train a new model, simply run,
python run.py
Once training is completed, the weights of the trained model are saved in saved-model/
directory.
To perform inference, simply run,
python inference.py <file>
where <file>
is the path of the numpy file you would like to perform an inference on.