This repository implements the algorithms described in our paper TransPCC: Towards Deep Point Cloud Compression via Transformers.
Install nvida-docker and follow these instructions
You can download the dataset from here povided by depoco and link the dataset to the docker container by configuring the Makefile
DATASETS=<path-to-your-data>
For building the Docker Container simply run
make build
in the root directory.
The first step is to run the docker container:
make run
The following commands assume to be run inside the docker container.
For training a network we first have to create the config file with all the parameters.
An example of this can be found in /depoco/config/transPCC_demo.yaml
.
Make sure to give each config file a unique experiment_id: ...
to not override previous models.
To train the network simply run
python3 transPCC_trainer.py -cfg <path-to-your-config>
Evaluating the network on the test set can be done by:
python3 evaluate.py -cfg <path-to-your-config>
All results will be saved in a dictonary.
If you find this paper helps your research, please kindly consider citing our paper in your publications.
@inproceedings{liang2022transpcc,
title={TransPCC: Towards Deep Point Cloud Compression via Transformers},
author={Liang, Zujie and Liang, Fan},
booktitle={Proceedings of the 2022 International Conference on Multimedia Retrieval (ICMR)},
year={2022}
}
This repo contains code modified from depoco, Many thanks for their efforts.