Prompt-DAS

The official code repository for the Miccai 2025 accepted paper:

Prompt-DAS: Annotation-Efficient Prompt Learning for Domain Adaptive Semantic Segmentation of Electron Microscopy Images

Requirements

To install requirements:

pip install -r requirements.txt

Datasets：

Download MitoEM dataset（MitoEM-Human and MitoEM-Rat）https://mitoem.grand-challenge.org/

Organize the dataset into the following directory structure：

dataset/
├── H/            
│   └─ test/              
│       ├── img
│       │    ├── im0400.png
│       │    └── ....
│       └── lab
│            ├── im0400.png
│            └── ....
└── R/                  
    └─ test/
        ├── img
        │    ├── im0400.png
        │    └── ....
        └── lab
             ├── im0400.png
             └── ....

Download Checkpoint

Click the links below to download the checkpoint for the corresponding dataset.

R-H.pth

H-R.pth

Training

Todo

Evaluation

To evaluate my model on R/H, run:

python Evaluation.py