DES-SAM

Distillation-Enhanced Semantic SAM for Cervical Nuclear Segmentation with Box Annotation

[Model] [Paper] [BibTeX]

Lina Huang, Yixiong Liang, JianFeng Liu

Model Overview

Overview of DES-SAM model architecture

Install

On an NVIDIA 3090 Tensor Core GPU machine, with CUDA toolkit enabled.

1. Download our repository and open the DES-SAM

git clone [email protected]:CVIU-CSU/DES-SAM.git
cd DES-SAM

2. Install MMDetection 🛠️Installation and its dependencies

# Step 1. Create a conda environment and activate it 
conda create --name dessam python=3.8 -y
conda activate dessam
# Step 2. Install PyTorch following official instructions, e.g.
pip install torch==1.13.1+cu116 torchvision==0.14.1+cu116 torchaudio==0.13.1 --extra-index-url https://download.pytorch.org/whl/cu116
# Step 3. MMDetection Installation 
pip install -U openmim
mim install mmengine
mim install "mmcv==2.0.1"
cd mmdetection
pip install -v -e .
# Step 4. Package Installation
pip install -r requirements.txt
# Step 5. SAM Installation
pip install segment-anything

Results and Models

• Visual Result

• Model Download

The MMDetection based models can be accessed from Baiduyun.

Train & Test

Download the pretrained model to train DES-SAM.

Our code is based on coco datasets, datasets need to be converted to coco first.

# Train
CUDA_VISIBLE_DEVICES=0,1,2,3,4,5,6,7 bash ./tools/dist_train.sh {path}/mmdetection/configs/_des_sam_/PatchSeg/des-sam-patch.py 8 

# Test
CUDA_VISIBLE_DEVICES=0,1,2,3,4,5,6,7 bash ./tools/dist_test.sh {path}/mmdetection/configs/_des_sam_/PatchSeg/des-sam-patch.py {model_path} 8

Acknowledgements

We would like to express our gratitude to the authors and developers of the exceptional repositories that this project is built upon:

• MMDETECTION
• CNSeg
• SAM
• VitDet
• Faster R-CNN
• BoxSnake
• BoxLevelSet

Their contributions have been invaluable to our work.

Citation

If you find it useful for your your research and applications, please cite using this BibTeX:

@inproceedings{huang2024des-sam,
  title={DES-SAM: Distillation-Enhanced Semantic SAM for Cervical Nuclear Segmentation with Box Annotation},
  author={Lina Huang, Yixiong Liang and Jianfeng Liu},
  booktitle={International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI)},
  year={2024},
  publisher={Springer}
}