Skip to content

Latest commit

 

History

History

WSL

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
configs
configs
 
 
scripts
scripts
 
 
tools
tools
 
 
wsl
wsl
 
 
.gitignore
.gitignore
 
 
README.md
README.md
 
 
requirements.txt
requirements.txt
 
 
setup.py
setup.py
 
 

README.md

Enabling Deep Residual Networks for Weakly Supervised Object Detection

By Yunhang Shen, Rongrong Ji, Yan Wang, Zhiwei Chen, Feng Zheng, Feiyue Huang, Yunsheng Wu.

ECCV 2020 Paper.

This project is based on Detectron2.

License

DRN-WSOD is released under the Apache 2.0 license.

Citing DRN-WSOD

If you find DRN-WSOD useful in your research, please consider citing:

@inproceedings{DRN-WSOD_2020_ECCV,
	author = {Shen, Yunhang and Ji, Rongrong and Wang, Yan and Chen, Zhiwei and Zheng, Feng and Huang, Feiyue and Wu, Yunsheng},
	title = {Enabling Deep Residual Networks for Weakly Supervised Object Detection},
	booktitle = {European Conference on Computer Vision (ECCV)},
	year = {2020},
}

Installation

Install our forked Detectron2:

git clone https://github.com/shenyunhang/DRN-WSOD-pytorch.git
cd DRN-WSOD-pytorch
python3 -m pip install -e .

If you have problem of installing Detectron2, please checking the instructions.

Install DRN-WSOD project:

cd projects/WSL
pip3 install -r requirements.txt
git submodule update --init --recursive
python3 -m pip install -e .
cd ../../

Dataset Preparation

Please follow this to creating symlinks for PASCAL VOC.

Download MCG proposal from here to detectron/datasets/data, and transform it to pickle serialization format:

cd datasets/proposals
tar xvzf MCG-Pascal-Main_trainvaltest_2007-boxes.tgz
cd ../../
python3 projects/WSL/tools/proposal_convert.py voc_2007_train datasets/proposals/MCG-Pascal-Main_trainvaltest_2007-boxes datasets/proposals/mcg_voc_2007_train_d2.pkl
python3 projects/WSL/tools/proposal_convert.py voc_2007_val datasets/proposals/MCG-Pascal-Main_trainvaltest_2007-boxes datasets/proposals/mcg_voc_2007_val_d2.pkl
python3 projects/WSL/tools/proposal_convert.py voc_2007_test datasets/proposals/MCG-Pascal-Main_trainvaltest_2007-boxes datasets/proposals/mcg_voc_2007_test_d2.pkl

Model Preparation

Download models from this here:

mv models $DRN-WSOD

Then we have the following directory structure:

DRN-WSOD
|_ models
|  |_ DRN-WSOD
|     |_ resnet18_ws_model_120.pkl
|     |_ resnet150_ws_model_120.pkl
|     |_ resnet101_ws_model_120.pkl
|_ ...

Quick Start: Using DRN-WSOD

WSDDN

ResNet18-WS

python3 projects/WSL/tools/train_net.py --num-gpus 4 --config-file projects/WSL/configs/PascalVOC-Detection/wsddn_WSR_18_DC5_1x.yaml OUTPUT_DIR output/wsddn_WSR_18_DC5_VOC07_`date +'%Y-%m-%d_%H-%M-%S'`

ResNet50-WS

python3 projects/WSL/tools/train_net.py --num-gpus 4 --config-file projects/WSL/configs/PascalVOC-Detection/wsddn_WSR_50_DC5_1x.yaml OUTPUT_DIR output/wsddn_WSR_50_DC5_VOC07_`date +'%Y-%m-%d_%H-%M-%S'`

ResNet101-WS

python3 projects/WSL/tools/train_net.py --num-gpus 4 --config-file projects/WSL/configs/PascalVOC-Detection/wsddn_WSR_101_DC5_1x.yaml OUTPUT_DIR output/wsddn_WSR_101_DC5_VOC07_`date +'%Y-%m-%d_%H-%M-%S'`

VGG16

python3 projects/WSL/tools/train_net.py --num-gpus 4 --config-file projects/WSL/configs/PascalVOC-Detection/wsddn_V_16_DC5_1x.yaml OUTPUT_DIR output/wsddn_V_16_DC5_VOC07_`date +'%Y-%m-%d_%H-%M-%S'`

OICR

ResNet18-WS

python3 projects/WSL/tools/train_net.py --num-gpus 4 --config-file projects/WSL/configs/PascalVOC-Detection/oicr_WSR_18_DC5_1x.yaml OUTPUT_DIR output/oicr_WSR_18_DC5_VOC07_`date +'%Y-%m-%d_%H-%M-%S'`

ResNet50-WS

python3 projects/WSL/tools/train_net.py --num-gpus 4 --config-file projects/WSL/configs/PascalVOC-Detection/oicr_WSR_50_DC5_1x.yaml OUTPUT_DIR output/oicr_WSR_50_DC5_VOC07_`date +'%Y-%m-%d_%H-%M-%S'`

ResNet101-WS

python3 projects/WSL/tools/train_net.py --num-gpus 4 --config-file projects/WSL/configs/PascalVOC-Detection/oicr_WSR_101_DC5_1x.yaml OUTPUT_DIR output/oicr_WSR_101_DC5_VOC07_`date +'%Y-%m-%d_%H-%M-%S'`

VGG16

python3 projects/WSL/tools/train_net.py --num-gpus 4 --config-file projects/WSL/configs/PascalVOC-Detection/oicr_V_16_DC5_1x.yaml OUTPUT_DIR output/oicr_V_16_DC5_VOC07_`date +'%Y-%m-%d_%H-%M-%S'`

OICR + Box Regression

ResNet18-WS

python3 projects/WSL/tools/train_net.py --num-gpus 4 --config-file projects/WSL/configs/PascalVOC-Detection/reg/oicr_WSR_18_DC5_1x.yaml OUTPUT_DIR output/oicr_reg_WSR_18_DC5_VOC07_`date +'%Y-%m-%d_%H-%M-%S'`

ResNet50-WS

python3 projects/WSL/tools/train_net.py --num-gpus 4 --config-file projects/WSL/configs/PascalVOC-Detection/reg/oicr_WSR_50_DC5_1x.yaml OUTPUT_DIR output/oicr_reg_WSR_50_DC5_VOC07_`date +'%Y-%m-%d_%H-%M-%S'`

ResNet101-WS

python3 projects/WSL/tools/train_net.py --num-gpus 4 --config-file projects/WSL/configs/PascalVOC-Detection/reg/oicr_WSR_101_DC5_1x.yaml OUTPUT_DIR output/oicr_reg_WSR_101_DC5_VOC07_`date +'%Y-%m-%d_%H-%M-%S'`

VGG16

python3 projects/WSL/tools/train_net.py --num-gpus 4 --config-file projects/WSL/configs/PascalVOC-Detection/reg/oicr_V_16_DC5_1x.yaml OUTPUT_DIR output/oicr_reg_V_16_DC5_VOC07_`date +'%Y-%m-%d_%H-%M-%S'`

PCL

ResNet18-WS

python3 projects/WSL/tools/train_net.py --num-gpus 4 --config-file projects/WSL/configs/PascalVOC-Detection/pcl_WSR_18_DC5_1x.yaml OUTPUT_DIR output/pcl_WSR_18_DC5_VOC07_`date +'%Y-%m-%d_%H-%M-%S'`

ResNet50-WS

python3 projects/WSL/tools/train_net.py --num-gpus 4 --config-file projects/WSL/configs/PascalVOC-Detection/pcl_WSR_50_DC5_1x.yaml OUTPUT_DIR output/pcl_WSR_50_DC5_VOC07_`date +'%Y-%m-%d_%H-%M-%S'`

ResNet101-WS

python3 projects/WSL/tools/train_net.py --num-gpus 4 --config-file projects/WSL/configs/PascalVOC-Detection/pcl_WSR_101_DC5_1x.yaml OUTPUT_DIR output/pcl_WSR_101_DC5_VOC07_`date +'%Y-%m-%d_%H-%M-%S'`

VGG16

python3 projects/WSL/tools/train_net.py --num-gpus 4 --config-file projects/WSL/configs/PascalVOC-Detection/pcl_V_16_DC5_1x.yaml OUTPUT_DIR output/pcl_V_16_DC5_VOC07_`date +'%Y-%m-%d_%H-%M-%S'`