DFENet

Deep Fourier-embedded Network for Bi-modal Salient Object Detection [paper (an initial version)]

• April 29, 2024
  The paper is undergoing peer review. The code will be released upon acceptance of the paper.
• 
• In this project, we proposed the deep Fourier-embedded network (DFENet), a purely Fourier-based model aimed at solving the high-resolution bi-modal inputs and feature fusion while minimizing memory consumption of GPU, outperforming existing state-of-the-art bi-modal salient object detection (BSOD) models on four RGB-T BSOD benchmark datasets. To the best of our knowledge, this is the first Fourier-based supervised model in a series of salient object detection tasks.
• Please cite our paper if you find it useful for your research.

@article{lyu2024deep,
  title={Deep Fourier-embedded Network for Bi-modal Salient Object Detection},
  author={Lyu, Pengfei and Yu, Xiaosheng and Wu, Chengdong and Rajapakse, Jagath C},
  journal={arXiv preprint arXiv:2411.18409},
  year={2024}
}

Requirements

List of prerequisites or required libraries for the project to run:

• Pytorch 2.0.0
• Cuda 11.8
• Python 3.8 or higher
• tensorboardX
• opencv-python
• timm == 0.5.4
• thop
• numpy

Datasets

• We conducted experiments to evaluate our DFENet on the VT821, VT1000, VT5000, and VI-RGBT1500 datasets. Please click for the corresponding dataset.

Pre-trained Weights of DFENet

Resolution	Backbone	weights
384 x 384	CDFFormer-m36	Link
512 x 512	CDFFormer-m36	Link

Results

• The results of our DFENet can be found at link.
• 

Evaluation Metrics Toolbox

• The Evaluation Metrics Toolbox is available here: link.

Acknowledgements

• Thanks to all the seniors and projects (e.g., DFFormer, FFL, UHDFour, MGAI, and SwinNet).

Contact Us

If you have any questions, please contact us ([email protected]).