Efficient Fourier Filtering Network with Contrastive Learning for AAV-based Unaligned Bi-modal Salient Object Detection [arXiv] [IEEE]
- In this project, we designed AlignSal, which achieves both real-time performance and high accuracy for AAV-based unaligned Bi-modal Salient Object Detection (BSOD).
- Please cite our paper if you find it useful for your research.
@ARTICLE{10975009,
author={Lyu, Pengfei and Yeung, Pak-Hei and Yu, Xiaosheng and Cheng, Xiufei and Wu, Chengdong and Rajapakse, Jagath C.},
journal={IEEE Transactions on Geoscience and Remote Sensing},
title={Efficient Fourier Filtering Network With Contrastive Learning for AAV-Based Unaligned Bimodal Salient Object Detection},
year={2025},
volume={63},
number={},
pages={1-12},
publisher={IEEE}
}
@ARTICLE{10982382,
author={Lyu, Pengfei and Yu, Xiaosheng and Chi, Jianning and Wu, Hao and Wu, Chengdong and Rajapakse, Jagath C.},
journal={IEEE Transactions on Image Processing},
title={TwinsTNet: Broad-View Twins Transformer Network for Bi-Modal Salient Object Detection},
year={2025},
volume={34},
number={},
pages={2796-2810},
publisher={IEEE}
}
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==0.6.13
- thop
- numpy
The results of our AlignSal and other SOTA models
- The Evaluation Metrics Toolbox is available here: link.
- Thanks to all the seniors, and projects (e.g., MROS, ContrastAlign, DCNet, and SwinNet).
If you have any questions, please contact us ([email protected]).