PointRCNN

3D Object Proposal Generation and Detection from Point Cloud [Multiclass]

Reference paper PointRCNN:3D Object Proposal Generation and Detection from Point Cloud, CVPR 2019.

Authors: Shaoshuai Shi, Xiaogang Wang, Hongsheng Li.

[arXiv]  [Project Page] 

Installation

Requirements

All the codes are tested in the following environment:

• Linux (tested on Ubuntu 14.04/16.04)
• Python 3.6+
• PyTorch 1.0

Install PointRCNN

a. Clone the PointRCNN_multiclass repository.

git clone https://github.com/sshaoshuai/PointRCNN.git

b. Install the dependent python libraries like easydict,tqdm, tensorboardX etc.

conda env create --file environment.yaml
conda activate pointrcnn

c. Build and install the pointnet2_lib, iou3d, roipool3d libraries by executing the following command:

sh build_and_install.sh

Dataset preparation

Please download the official KITTI 3D object detection dataset and organize the downloaded files as follows:

PointRCNN
├── data
│   ├── KITTI
│   │   ├── ImageSets
│   │   ├── object
│   │   │   ├──training
│   │   │      ├──calib & velodyne & label_2 & image_2 & (optional: planes)
│   │   │   ├──testing
│   │   │      ├──calib & velodyne & image_2
├── lib
├── pointnet2_lib
├── tools

Here the images are only used for visualization and the road planes are optional for data augmentation in the training.

Training

a. Create ground truth Multiclass File (Need to change tools/default.yaml CLASSES: Car -> Multiclass)

python generate_gt_database.py --class_name 'Multiclass' --split train

b. RPN training (Create proposal candidates)

python train_rcnn.py --cfg_file cfgs/default.yaml --batch_size 16 --train_mode rpn --epochs 200

c. RCNN training (Create target proposals)

python train_rcnn.py --cfg_file cfgs/default.yaml --batch_size 4 --train_mode rcnn --epochs 100 --ckpt_save_interval 2 --rpn_ckpt ../output/rpn/default/ckpt/checkpoint_epoch_200.pth

Inference

python eval_rcnn.py --cfg_file cfgs/default.yaml --ckpt ../output/rcnn/default/ckpt/checkpoint_epoch_100.pth --eval_mode rcnn --test

If you want to test, not eval, change TEST.SPLIT: val => TEST.SPLIT: test in tools/default.yaml general training config part

Citation

If you find this work useful in your research, please consider cite:

@InProceedings{Shi_2019_CVPR,
    author = {Shi, Shaoshuai and Wang, Xiaogang and Li, Hongsheng},
    title = {PointRCNN: 3D Object Proposal Generation and Detection From Point Cloud},
    booktitle = {The IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
    month = {June},
    year = {2019}
}

To Do List

• Implement multiclass one shot learning and Release
• Increase Pedestrain, Cyclist Detection Accuracy (Data unbalance problem)
• Annotation arrangement