/GLNet_TCYB2022

The results and code of our IEEE TCYB 2022 paper, titled "Global-and-Local Collaborative Learning for Co-Salient Object Detection"

GLNet_TCYB2022

Runmin Cong, Ning Yang, Chongyi Li, Huazhu Fu, Yao Zhao, Qingming Huang, and Sam Kwong, Global-and-local collaborative learning for co-Salient object detection, IEEE Transactions on Cybernetics, 2022.

Project: https://rmcong.github.io/proj_GLNet.html

Results of GLNet:

  • Results:
    • We provide the resutls of our GLNet on Cosal2015, iCoseg, and MSRC.
Baidu Cloud: https://pan.baidu.com/s/1sXBc4H3fKK8Y8ceaU4AjSQ  Password: 0224

Pytorch Code of GLNet:

  • Pytorch implementation of GLNet
  • Pretrained model:
    • We provide our testing code. If you test our model, please download the pretrained model, unzip it, and put the checkpoint model_GLNet.pth to Checkpoints/trained/ folder and put the pretrained backbone backbone_v.pth to Checkpoints/warehouse/ folder.
    • Pretrained model download:
Baidu Cloud: https://pan.baidu.com/s/1sXBc4H3fKK8Y8ceaU4AjSQ  Password: 0224

Requirements

  • Python 3.7
  • Pytorch 1.5.1
  • torchvision

Data Preprocessing

  • We resize the images of original test datasets. Please download the resized data, and put the data to Data/ folder.
  • Resized test datasets:
Baidu Cloud: https://pan.baidu.com/s/1sXBc4H3fKK8Y8ceaU4AjSQ  Password: 0224

Test

python test.py
  • You can find the results in the 'Outputs/' folder.

If you use our GLNet, please cite our paper:

@article{GLNet,
    title={Global-and-local collaborative learning for co-Salient object detection},
    author={Cong, Runmin and Yang, Ning and Li, Chongyi and Fu, Huazhu and Zhao, Yao and Huang, Qingming and Kwong, Sam},
    journal={IEEE Trans. Cybern.},
    year={early access, doi: 10.1109/TCYB.2022.3169431},
    publisher={IEEE}
}

Contact Us:

If you have any questions, please contact Runmin Cong (rmcong@bjtu.edu.cn) or Ning Yang (ningyang@bjtu.edu.cn).