DifIISR: Diffusion Model with Gradient Guidance for Infrared Image Super-Resolution

DifIISR: Diffusion Model with Gradient Guidance for Infrared Image Super-Resolution [CVPR 2025]

Updates

[2025-3-4] You can find our paper here.
[2025-2-27] Our paper has been accepted by CVPR 2025, and the code will be released soon.

Environment

# create virtual environment
conda create -n DifIISR python=3.10
conda activate DifIISR
# install requirements
pip install -r requirements.txt

Test

Our checkpoints can be found in Google drive, put it in 'DifIISR/weights/', you can test our method through

CUDA_VISIBLE_DEVICES=0 python inference.py -input dataset/test/LR -output results -reference dataset/test/HR --config configs/DifIISR_test.yaml

Citation

@article{li2025difiisr,
  title={DifIISR: A Diffusion Model with Gradient Guidance for Infrared Image Super-Resolution},
  author={Li, Xingyuan and Wang, Zirui and Zou, Yang and Chen, Zhixin and Ma, Jun and Jiang, Zhiying and Ma, Long and Liu, Jinyuan},
  journal={arXiv preprint arXiv:2503.01187},
  year={2025}
}

Contact

If you have any questions, feel free to contact me through ziruiwang0625@gmail.com。

Acknowledgement

Our codes are based on ResShift, SinSR, thanks for their contribution.