# L-SAHGNet

**Repository Path**: RobVIP-Lab/l-sahgnet

## Basic Information

- **Project Name**: L-SAHGNet
- **Description**: The codes for the work "Attention-Aware Laparoscopic Image Desmoking Network with Lightness Embedding and Hybrid Guided Embedding "(https://arxiv.org/abs/2404.07556).  Published in IEEE ROBIO 2024.
- **Primary Language**: Unknown
- **License**: Not specified
- **Default Branch**: main
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2024-06-19
- **Last Updated**: 2025-09-05

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

#  L-SAHGNet
The codes for the work "Attention-Aware Laparoscopic Image Desmoking Network with Lightness Embedding and Hybrid Guided Embedding "(https://arxiv.org/abs/2404.07556) and "Smoke Attention Based Laparoscopic Image Desmoking Network with Hybrid Guided Embedding"(ROBIO 2024, https://ieeexplore.ieee.org/document/10907529). The code is based on the [Swin-Unet](https://github.com/HuCaoFighting/Swin-Unet).

## 1. Download pre-trained model and data

- Please go to ["./datasets/README.md"](datasets/README.md) for details.

## 2. Environment

- Please prepare an environment with python=3.9, and then use the command "pip install -r requirements.txt" for the dependencies.

## 3. Train/Test

- Train

```bash
python train_MIS.py --dataset PSv2rs --cfg configs/deswin_patch4_windows8_256.yaml --root_path your DATA_DIR --max_epochs 300 -output_dir your OUT_DIR  --img_size 256 --base_lr 0.0001 --batch_size 24
```

- Test 

```bash
python test.py --dataset PSv2rs --cfg configs/deswin_patch4_windows8_256.yaml --dataset_path  F:\render\PureSmoke\PSv2rs --output_dir ./outtest --max_epoch 150 --base_lr 0.05 --img_size 256 --batch_size 1
```

## References
* [Swin-Unet](https://github.com/HuCaoFighting/Swin-Unet)

## Citation

```bibtex
@misc{liu2024attentionaware,
      title={Attention-Aware Laparoscopic Image Desmoking Network with Lightness Embedding and Hybrid Guided Embedding}, 
      author={Ziteng Liu and Jiahua Zhu and Bainan Liu and Hao Liu and Wenpeng Gao and Yili Fu},
      year={2024},
      eprint={2404.07556},
      archivePrefix={arXiv},
      primaryClass={id='eess.IV' full_name='Image and Video Processing' is_active=True alt_name=None in_archive='eess' is_general=False description='Theory, algorithms, and architectures for the formation, capture, processing, communication, analysis, and display of images, video, and multidimensional signals in a wide variety of applications. Topics of interest include: mathematical, statistical, and perceptual image and video modeling and representation; linear and nonlinear filtering, de-blurring, enhancement, restoration, and reconstruction from degraded, low-resolution or tomographic data; lossless and lossy compression and coding; segmentation, alignment, and recognition; image rendering, visualization, and printing; computational imaging, including ultrasound, tomographic and magnetic resonance imaging; and image and video analysis, synthesis, storage, search and retrieval.'}
}
```

```bibtex
@INPROCEEDINGS{liu2024Smokeattention,
  author={Liu, Ziteng and Gao, Wenpeng and Zhu, Jiahua and Liu, Bainan and Fu, Yili},
  booktitle={2024 IEEE International Conference on Robotics and Biomimetics (ROBIO)}, 
  title={Smoke Attention Based Laparoscopic Image Desmoking Network with Hybrid Guided Embedding}, 
  year={2024},
  volume={},
  number={},
  pages={1018-1023},
  keywords={Laparoscopes;Training;PSNR;Minimally invasive surgery;Atmospheric modeling;Scattering;Real-time systems;Robots;Videos;Testing},
  doi={10.1109/ROBIO64047.2024.10907529}}

```