SFIN

Code and data of CVPR 2025 paper "Noise Calibration and Spatial-Frequency Interactive Network for STEM Image Enhancement". Paper PDF

How to Use

1. Down checkpoints here. Files with _bf are for BF mode and others are for HAADF mode.
2. Unzip test dataset haadf_data_test and bf_data_test.zip.
3. Run ours_gpu_demo.py to get enhancement/detection results.
4. Run metrics.py to get PSNR/SSIM metrics.

If you have any other needs, please contact us via email [email protected] or my WeChat: linsfriend.

Citation

If you find the code helpful in your resarch or work, please cite the following paper(s).

@article{Li2025SFIN,
    title = {Noise Calibration and Spatial-Frequency Interactive Network for STEM Image Enhancement},
    author = {Li, Hesong and Wu, Ziqi and Shao, Ruiwen and Zhang, Tao and Fu, Ying},
    booktitle = {Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition},
    year = {2025},
}