PillarNeXt: Rethinking Network Designs for 3D Object Detection in LiDAR Point Clouds

Jinyu Li, Chenxu Luo, Xiaodong Yang <br>
PillarNeXt: Rethinking Network Designs for 3D Object Detection in LiDAR Point Clouds, CVPR 2023 <br>
[Paper] [Poster]

<p align="left"> <img src='docs/teaser_figure.png' height="410px"/> </p>

Get Started

Installation

Please refer to INSTALL to set up environment and install dependencies (see detail in Dockerfile).

Data Preparation

Please follow the instructions in DATA.

Training and Evaluation

Please follow the instructions in RUN.

Main Results

nuScenes (Val)

Model	mAP	NDS	Checkpoint
PillarNeXt-B	62.5	68.8	[Google Drive] [Baidu Cloud]

Waymo Open Dataset

Split	#Frames	Veh L2 3D APH	Ped L2 3D APH	Cyc L2 3D APH
Val	1	69.8	69.8	69.6
Val	3	72.4	75.2	75.7
Test	3	75.8	76.0	70.6

Citation

Please cite the following paper if this repo helps your research:

@inproceedings{li2023pillarnext,
  title={PillarNeXt: Rethinking Network Designs for 3D Object Detection in LiDAR Point Clouds},
  author={Li, Jinyu and Luo, Chenxu and Yang, Xiaodong},
  booktitle={IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
  year={2023}
}

Acknowledgement

We thank the authors for the multiple great open-sourced repos, including Det3D, CenterPoint and OpenPCDet.

License

Copyright (C) 2023 QCraft. All rights reserved. Licensed under the CC BY-NC-SA 4.0 (Attribution-NonCommercial-ShareAlike 4.0 International). The code is released for academic research use only. For commercial use, please contact business@qcraft.ai.