X-View: Non-Egocentric Multi-View 3D Object Detector
Authors: Liang Xie, Guodong Xu, Deng Cai, Xiaofei He
Introduction: This repository provides an implementation for our paper "X-View: Non-Egocentric Multi-View 3D Object Detector" published on T-IP. This repository is based on OpenPCDet. X-View generalizes the research on 3D multi-view learning. Specifically, X-view breaks through the traditional limitation about the perspective view whose original point must be consistent with the 3D Cartesian coordinate to overcome the drawbacks of the multi-view methods.
The environment requirements are same as the original code base and you can refer the instructions to install the environments and prepare the datasets.
Please refer the instructions in the original code base to perform training and testing. The only thing you should do is changing the path to the configure file.
For example,
sh scripts/dist_train.sh 4 --cfg_file tools/cfgs/kitti_models/second_mv2.yaml
Please consider citing X-View in your publications if it helps your research. :)
@ARTICLE{10049775,
author={Xie, Liang and Xu, Guodong and Cai, Deng and He, Xiaofei},
journal={IEEE Transactions on Image Processing},
title={X-View: Non-Egocentric Multi-View 3D Object Detector},
year={2023},
volume={32},
number={},
pages={1488-1497},
doi={10.1109/TIP.2023.3245337}
}If you have any questions about this work, feel free to contact us through email (Liang Xie: lilydedbb@gmail.com) or Github issues.