This is an official pytorch implementation of the IROS 2022 paper CA-SpaceNet: Counterfactual Analysis for 6D Pose Estimation in Space. In this repository, we provide PyTorch code for training and testing the proposed CA-SpaceNet on the Swisscube and SPEED dataset.
[arXiv][supp video]
If this repository is helpful to you, please star it. If you find our work useful in your research, please consider citing:
@article{CA-SpaceNet2022,
title={CA-SpaceNet: Counterfactual Analysis for 6D Pose Estimation in Space},
author={Wang, Shunli and Wang, Shuaibing and Jiao, Bo and Yang, Dingkang and Su, Liuzhen and Zhai, Peng and Chen, Chixiao and Zhang, Lihua},
journal={2022 IEEE/RSJ International Conference on Intelligent Robots and Systems},
year={2022}
}In this repository, we provide two separate folders to store two sets of codes on the Swisscube dataset and the SPEED dataset, respectively.
-
./CA-SpaceNet-on-Swisscube/: This folder contains all codes of the proposed CA-SpaceNet on the Swisscube dataset and pre-trained models. Please refer to Swisscube_README for more details.
-
./CA-SpaceNet-on-SPEED/: This folder contains all codes of the proposed CA-SpaceNet on the SPEED dataset and pre-trained models. Please refer to SPEED_README for more details.
Some of the code is borrowed from the Swisscube project. We are very grateful for their wonderful implementation.
If you have any questions about our work, please contact slwang19@fudan.edu.cn or sbwang21@m.fudan.edu.cn.