Skip to content

mingsiliu557/DiffCTA

Repository files navigation

Leveraging Diffusion Models for Continual Test-Time Adaptation in Fundus Image Classification

📌 Updates

🗓 2025.03.19: Upload the code for Glaucoma&Diabetic classification.
🗓 2025.03.06: Repository created.

✅ TODO

  • Extend the experiment on segmentation task.
  • Code will be released soon.

🛠️ Dependencies & Installation

1️⃣ Clone the Repository

git clone git@github.com:anonymous/DiffCTA.git
cd DiffCTA

2️⃣ Create Conda Environment & Install Dependencies

conda create -n DiffCTA python=3.8 -y  
conda activate DiffCTA 
pip3 install -r requirements.txt  

🚀 Get Started

📂 Dataset and Checkpoint Preparation

  • Download the dataset and Checkpoint using the following command:
wget https://oneflow-static.oss-cn-beijing.aliyuncs.com/data_lx/Fundus.zip

Generate Adapted Image

bash optic_adapt.sh

⚡ Quick Test 🏂


- Run the following command to perform a quick inference:  
```bash
bash TTA.sh

📊 Results

Method Domain A Domain B Domain C Domain D Domain E AVG
Source Only 68.37 50.68 65.74 34.98 43.43 52.64
TENT 65.84 58.84 60.36 28.42 42.76 51.24
CoTTA 64.01 58.51 61.25 24.02 33.59 48.28
EATA 66.46 58.50 63.34 33.41 40.42 52.43
SAR 66.57 58.81 63.21 32.87 33.52 51.00
DDA 69.71 56.06 67.20 34.22 39.73 53.38
DiffCTA 70.47 59.34 68.46 35.59 45.93 55.96

📜 Citation (TODO)

📄 License

The code and models are licensed under MIT License.

📬 Contact (anonymous)

🙌 Acknowledgement

The code is inspired by VPTTA, DDA

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published