Code and data for CVPR26 paper "Statistical Characteristic-Guided Denoising for Rapid High-Resolution Transmission Electron Microscopy Imaging" Paper PDF

1. Unzip tem_test_data4.zip. This is the test dataset.
2. Download and unzip tem_data4.zip from Baidu disk or GoogleDrive. This is the training dataset.
3. Run train_convLast_std_tem_data4.py to train and test our SCGN. convLast_std.py is the SCGN network architecture.
4. Visual results will be saved in "convLast_std_tem_data4_result" folder. PSNR, SSIM, and IOU results will be printed.
5. convLast_std_tem_data4_100.pth is our obtained weight. This will be loaded automatically to test. If you want to retrain, you can delete this weight.

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

@article{SCGN,
    title = {Statistical Characteristic-Guided Denoising for Rapid High-Resolution Transmission Electron Microscopy Imaging},
    author = {Li, Hesong and Wu, Ziqi and Shao, Ruiwen and Fu, Ying},
    booktitle = {Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition},
    year = {2026},
}