A deep learning method for pointcloud object detection.
This is an anchor free method for pointcloud object detection by using bird eye view.
This project is not finished yet, it has a lot of parts to be improved.
If you are intreseted in this project, you can try to change the code and make this work better.
git clone https://github.com/wangx1996/CenterBEV.git CenterBEV
cd CenterBEV/
pip install -r requirements.txt
for anaconda
conda install scikit-image scipy numba pillow matplotlib
pip install fire tensorboardX protobuf opencv-python
Please download DCNV2 from https://github.com/jinfagang/DCNv2_latest to fit torch 1.
Put the file into
./src/model/
then
./make.sh
export NUMBAPRO_CUDA_DRIVER=/usr/lib/x86_64-linux-gnu/libcuda.so
export NUMBAPRO_NVVM=/usr/local/cuda/nvvm/lib64/libnvvm.so
export NUMBAPRO_LIBDEVICE=/usr/local/cuda/nvvm/libdevice
KITTI dataset
You can Download the KITTI 3D object detection dataset from here.
It includes: Velodyne point clouds (29 GB)
Training labels of object data set (5 MB)
Camera calibration matrices of object data set (16 MB)
Left color images of object data set (12 GB)
Data structure like
└── KITTI_DATASET_ROOT
├── classes_names.txt
├── training <-- 7481 train data
| ├── image_2 <-- for visualization
| ├── calib
| ├── label_2
| └── velodyne
└── testing <-- 7580 test data
| ├── image_2 <-- for visualization
| ├── calib
| └── velodyne
└── ImageSets
├── train.txt
├── val.txt
├── trainval.txt
└── test.txt
First, make sure the dataset dir is right in your train.py file
Then run
python train.py --gpu_idx 0 --arch DLA_34 --saved_fn dla34 --batch_size 4
Tensorboard
cd logs/<saved_fn>/tensorboard/
tensorboard --logdir=./
if you want to test the work
python test.py --gpu_idx 0 --arch DLA_34 --pretrained_path ../checkpoints/**/** --peak_thresh 0.4
if you want to evaluate the work
python evaluate.py --gpu_idx 0 --arch DLA_34 --pretrained_path ../checkpoints/**/**
Checkpoints: https://drive.google.com/drive/folders/1YNgJ-DYt4AQYNqI8SysauMC1MRLG1vui?usp=sharing
Thanks for all the great works.
[1] SFA3D
[2] Complex_Yolo
[3] CenterNet: Objects as Points, [PyTorch Implementation]