PVNet: Pixel-wise Voting Network for 6DoF Pose Estimation

PVNet: Pixel-wise Voting Network for 6DoF Pose Estimation
Sida Peng, Yuan Liu, Qixing Huang, Xiaowei Zhou, Hujun Bao
CVPR 2019 oral
Project Page

Any questions or discussions are welcomed!

Introduction

Thanks Haotong Lin for providing the clean version of PVNet and reproducing the results.

The structure of this project is described in project_structure.md.

Installation

One way is to set up the environment with docker. See this.

Thanks Floris Gaisser for providing the docker implementation.

Another way is to use the following commands.

Set up the python environment:

conda create -n pvnet python=3.7
conda activate pvnet

# install torch 1.1 built from cuda 9.0
pip install torch==1.1.0 -f https://download.pytorch.org/whl/cu90/stable

pip install Cython==0.28.2
sudo apt-get install libglfw3-dev libglfw3
pip install -r requirements.txt

Compile cuda extensions under lib/csrc:

ROOT=/path/to/clean-pvnet
cd $ROOT/lib/csrc
export CUDA_HOME="/usr/local/cuda-9.0"
cd dcn_v2
python setup.py build_ext --inplace
cd ../ransac_voting
python setup.py build_ext --inplace
cd ../nn
python setup.py build_ext --inplace
cd ../fps
python setup.py build_ext --inplace

# If you want to use the uncertainty-driven PnP
cd ../uncertainty_pnp
sudo apt-get install libgoogle-glog-dev
sudo apt-get install libsuitesparse-dev
sudo apt-get install libatlas-base-dev
python setup.py build_ext --inplace

Set up datasets:

ROOT=/path/to/clean-pvnet
cd $ROOT/data
ln -s /path/to/linemod linemod
ln -s /path/to/linemod_orig linemod_orig
ln -s /path/to/occlusion_linemod occlusion_linemod

# the following is used for tless
ln -s /path/to/tless tless
ln -s /path/to/cache cache
ln -s /path/to/SUN2012pascalformat sun

Download datasets which are formatted for this project:

linemod
linemod_orig: The dataset includes the depth for each image.
occlusion linemod
truncation linemod: Check TRUNCATION_LINEMOD.md for the information about the Truncation LINEMOD dataset.
Tless: cat tlessa* | tar xvf - -C ..
Tless cache data: It is used for training and testing on Tless.
SUN2012pascalformat

Testing

Testing on Linemod

We provide the pretrained models of objects on Linemod, which can be found at here.

Take the testing on cat as an example.

Prepare the data related to cat:

python run.py --type linemod cls_type cat

Download the pretrained model of cat and put it to $ROOT/data/model/pvnet/cat/199.pth.

Test:

python run.py --type evaluate --cfg_file configs/linemod.yaml model cat cls_type cat
python run.py --type evaluate --cfg_file configs/linemod.yaml test.dataset LinemodOccTest model cat cls_type cat

Test with icp:

python run.py --type evaluate --cfg_file configs/linemod.yaml model cat cls_type cat test.icp True
python run.py --type evaluate --cfg_file configs/linemod.yaml test.dataset LinemodOccTest model cat cls_type cat test.icp True

Test with the uncertainty-driven PnP:

export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:./lib/csrc/uncertainty_pnp/lib
python run.py --type evaluate --cfg_file configs/linemod.yaml model cat cls_type cat test.un_pnp True
python run.py --type evaluate --cfg_file configs/linemod.yaml test.dataset LinemodOccTest model cat cls_type cat test.un_pnp True

Testing on Tless

We provide the pretrained models of objects on Tless, which can be found at here.

Download the pretrained models and put them to $ROOT/data/model/pvnet/.

Test:

python run.py --type evaluate --cfg_file configs/tless/tless_01.yaml
# or
python run.py --type evaluate --cfg_file configs/tless/tless_01.yaml test.vsd True

Visualization

Visualization on Linemod

Take the cat as an example.

Prepare the data related to cat:

python run.py --type linemod cls_type cat

Download the pretrained model of cat and put it to $ROOT/data/model/pvnet/cat/199.pth.

Visualize:

python run.py --type visualize --cfg_file configs/linemod.yaml model cat cls_type cat

If setup correctly, the output will look like

Visualize with a detector:

Download the pretrained models here and put them to $ROOT/data/model/pvnet/pvnet_cat/59.pth and $ROOT/data/model/ct/ct_cat/9.pth
```
python run.py --type detector_pvnet --cfg_file configs/ct_linemod.yaml
```

Visualization on Tless

Visualize:

python run.py --type visualize --cfg_file configs/tless/tless_01.yaml
# or
python run.py --type visualize --cfg_file configs/tless/tless_01.yaml test.det_gt True

Training

Training on Linemod

Prepare the data related to cat:

python run.py --type linemod cls_type cat

Train:

python train_net.py --cfg_file configs/linemod.yaml model mycat cls_type cat

The training parameters can be found in project_structure.md.

Training on Tless

Train:

python train_net.py --cfg_file configs/tless/tless_01.yaml

Tensorboard

tensorboard --logdir data/record/pvnet

If setup correctly, the output will look like

Training on the custom object

An example dataset can be downloaded at here.

Create a dataset using https://github.com/F2Wang/ObjectDatasetTools

Organize the dataset as the following structure:

├── /path/to/dataset
│   ├── model.ply
│   ├── camera.txt
│   ├── diameter.txt  // the object diameter, whose unit is meter
│   ├── rgb/
│   │   ├── 0.jpg
│   │   ├── ...
│   │   ├── 1234.jpg
│   │   ├── ...
│   ├── mask/
│   │   ├── 0.png
│   │   ├── ...
│   │   ├── 1234.png
│   │   ├── ...
│   ├── pose/
│   │   ├── pose0.npy
│   │   ├── ...
│   │   ├── pose1234.npy
│   │   ├── ...
│   │   └──

Create a soft link pointing to the dataset:

ln -s /path/to/custom_dataset data/custom

Process the dataset:
```
python run.py --type custom
```

Train:

python train_net.py --cfg_file configs/custom.yaml train.batch_size 4

Watch the training curve:
```
tensorboard --logdir data/record/pvnet
```

Visualize:

python run.py --type visualize --cfg_file configs/custom.yaml

Test:

python run.py --type evaluate --cfg_file configs/custom.yaml

An example dataset can be downloaded at here.

Citation

If you find this code useful for your research, please use the following BibTeX entry.

@inproceedings{peng2019pvnet,
  title={PVNet: Pixel-wise Voting Network for 6DoF Pose Estimation},
  author={Peng, Sida and Liu, Yuan and Huang, Qixing and Zhou, Xiaowei and Bao, Hujun},
  booktitle={CVPR},
  year={2019}
}

Acknowledgement

This work is affliated with ZJU-SenseTime Joint Lab of 3D Vision, and its intellectual property belongs to SenseTime Group Ltd.

Copyright (c) ZJU-SenseTime Joint Lab of 3D Vision. All Rights Reserved.

Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at

    http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.

yinglu-ecl / clean-pvnet

PVNet: Pixel-wise Voting Network for 6DoF Pose Estimation

Introduction

Installation

Testing

Testing on Linemod

Testing on Tless

Visualization

Visualization on Linemod

Visualization on Tless

Training

Training on Linemod

Training on Tless

Tensorboard

Training on the custom object

Citation

Acknowledgement

About

Languages