Deep learning-based cell segmentation using PyTorch and the MoNuSeg dataset.
The MoNuSeg dataset is used for training and evaluating segmentation models for cell images. This project includes functionalities for data loading, preprocessing, model training, and evaluation.
To set up the project, clone the repository and install the required packages:
git clone <repository-url>
cd cell-segmentation
pip install -r requirements.txt-
Data Loading: The dataset can be loaded using the
Datasetclass defined insrc/data/dataset.py. This class handles loading and preprocessing the MoNuSeg dataset, including image patching. -
Training the Model: To train the UNet model, run the following command:
python src/train.py- Data Exploration: Use the Jupyter notebook located in
notebooks/data_exploration.ipynbto visually inspect the dataset and explore its characteristics.
Configuration settings can be modified in src/config.py or configs/default.yaml to adjust paths and model parameters.
Evaluation metrics such as accuracy, precision, and recall can be calculated using the functions defined in src/utils/metrics.py.
Training progress and results can be visualized using the functions in src/utils/visualization.py.
This project is licensed under the MIT License.