Created by Charles R. Qi, Or Litany, Kaiming He and Leonidas Guibas from Facebook AI Research and Stanford University.
This repository is code release for ICCV 2019 paper (arXiv report here), and also the finetuning code for the paper DepthContrast.
In this repository, we provide VoteNet model implementation (with Pytorch) as well as data preparation, training, finetuning and evaluation scripts on SUN RGB-D and ScanNet.
If you find our work useful in your research, please consider citing:
@inproceedings{qi2019deep,
author = {Qi, Charles R and Litany, Or and He, Kaiming and Guibas, Leonidas J},
title = {Deep Hough Voting for 3D Object Detection in Point Clouds},
booktitle = {Proceedings of the IEEE International Conference on Computer Vision},
year = {2019}
}
Add our paper link here:
Install Pytorch and Tensorflow (for TensorBoard). It is required that you have access to GPUs. Matlab is required to prepare data for SUN RGB-D. The code is tested with Ubuntu 18.04, Pytorch v1.1, TensorFlow v1.14, CUDA 10.0 and cuDNN v7.4. Note: After a code update on 2/6/2020, the code is now also compatible with Pytorch v1.2+
Compile the CUDA layers for PointNet++, which we used in the backbone network:
cd pointnet2
python setup.py install
To see if the compilation is successful, try to run python models/votenet.py to see if a forward pass works.
Install the following Python dependencies (with pip install):
matplotlib
opencv-python
plyfile
'trimesh>=2.35.39,<2.35.40'
'networkx>=2.2,<2.3'
You can download pre-trained models and sample point clouds HERE.
Unzip the file under the project root path (/path/to/project/demo_files) and then run:
python demo.py
The demo uses a pre-trained model (on SUN RGB-D) to detect objects in a point cloud from an indoor room of a table and a few chairs (from SUN RGB-D val set). You can use 3D visualization software such as the MeshLab to open the dumped file under demo_files/sunrgbd_results to see the 3D detection output. Specifically, open ***_pc.ply and ***_pred_confident_nms_bbox.ply to see the input point cloud and predicted 3D bounding boxes.
You can also run the following command to use another pretrained model on a ScanNet:
python demo.py --dataset scannet --num_point 40000
Detection results will be dumped to demo_files/scannet_results.
For SUN RGB-D, follow the README under the sunrgbd folder.
For ScanNet, follow the README under the scannet folder.
For Matterport3D, follow the README under the mp3d folder.
For S3DIS, follow the README under the s3dis folder.
To train a new VoteNet model on SUN RGB-D data (depth images):
CUDA_VISIBLE_DEVICES=0 python train.py --dataset sunrgbd --log_dir log_sunrgbd
You can use CUDA_VISIBLE_DEVICES=0,1,2 to specify which GPU(s) to use. Without specifying CUDA devices, the training will use all the available GPUs and train with data parallel (Note that due to I/O load, training speedup is not linear to the nubmer of GPUs used). Run python train.py -h to see more training options (e.g. you can also set --model boxnet to train with the baseline BoxNet model).
While training you can check the log_sunrgbd/log_train.txt file on its progress, or use the TensorBoard to see loss curves.
For finetuning with our pretrained model:
CUDA_VISIBLE_DEVICES=0 python train.py --dataset sunrgbd --log_dir log_sunrgbd --no_height --pre_checkpoint_path /path/to/pretrained_model --batch_size 16
You can also specify the scan_idx for limited label training experiments and specify the scale parameter for 1x, 2x, 3x or 4x pointnet finetuning. The sampled indexes used in our experiments have been uploaded to sunrgbd/meta_data.
To test the trained model with its checkpoint:
python eval.py --dataset sunrgbd --checkpoint_path log_sunrgbd/checkpoint.tar --dump_dir eval_sunrgbd --cluster_sampling seed_fps --use_3d_nms --use_cls_nms --per_class_proposal
Example results will be dumped in the eval_sunrgbd folder (or any other folder you specify). You can run python eval.py -h to see the full options for evaluation. After the evaluation, you can use MeshLab to visualize the predicted votes and 3D bounding boxes (select wireframe mode to view the boxes).
Final evaluation results will be printed on screen and also written in the log_eval.txt file under the dump directory. In default we evaluate with both [email protected] and [email protected] with 3D IoU on oriented boxes. A properly trained VoteNet should have around 57 [email protected] and 32 [email protected].
To train a VoteNet model on Scannet data (fused scan):
CUDA_VISIBLE_DEVICES=0 python train.py --dataset scannet --log_dir log_scannet --num_point 40000
For finetuning with our pretrained model:
CUDA_VISIBLE_DEVICES=0 python train.py --dataset scannet --log_dir log_scannet --no_height --pre_checkpoint_path /path/to/pretrained_model --batch_size 8
You can also specify the scan_idx for limited label training experiments and specify the scale parameter for 1x, 2x, 3x or 4x pointnet finetuning. The sampled indexes used in our experiments have been uploaded to scannet/meta_data.
To test the trained model with its checkpoint:
python eval.py --dataset scannet --checkpoint_path log_scannet/checkpoint.tar --dump_dir eval_scannet --num_point 40000 --cluster_sampling seed_fps --use_3d_nms --use_cls_nms --per_class_proposal
Example results will be dumped in the eval_scannet folder (or any other folder you specify). In default we evaluate with both [email protected] and [email protected] with 3D IoU on axis aligned boxes. A properly trained VoteNet should have around 58 [email protected] and 35 [email protected].
Matterport3D and Stanford3D are very similar compared to ScanNet.
To train a VoteNet model on Matterport3D and Stanford3D (fused scan):
CUDA_VISIBLE_DEVICES=0 python train.py --dataset mp3d/s3dis --log_dir log
For finetuning with our pretrained model:
CUDA_VISIBLE_DEVICES=0 python train.py --dataset mp3d/s3dis --log_dir log --no_height --pre_checkpoint_path /path/to/pretrained_model --batch_size 8
You can also specify the scan_idx for limited label training experiments and specify the scale parameter for 1x, 2x, 3x or 4x pointnet finetuning. The sampled indexes used in our experiments have been uploaded to dataset/meta_data.
To test the trained model with its checkpoint, follow the samilar scripts compared to scannet.
[For Pro Users] If you have your own dataset with point clouds and annotated 3D bounding boxes, you can create a new dataset class and train VoteNet on your own data. To ease the proces, some tips are provided in this doc.
votenet is relased under the MIT License. See the LICENSE file for more details.
10/20/2019: Fixed a bug of the 3D interpolation customized ops (corrected gradient computation). Re-training the model after the fix slightly improves mAP (less than 1 point).