This project contains the code for training and evaluating a deep learning model designed for triage in the PAD-UFES-20+ dataset. It utilizes the raug library for model training and evaluation.

• benchmarks/: Contains scripts for running experiments.
  • kfold.py: Main script for running k-fold cross-validation experiments.
  • train.py: Contains the training and evaluation logic.
  • pad20plus/: Dataset-specific configurations and preprocessing scripts.
• models/: Contains model definitions and the model hub.
  • models_hub.py: Factory for creating model instances.
  • mobilenet.py: Implementation of MobileNet models.
  • metablock.py: Implementation of the MetaBlock for feature fusion.
• config.py: Configuration file for dataset paths.
• raug/: Submodule containing the training and evaluation framework.

1. Clone the repository and submodules:

# Initialize the raug submodule
git submodule update --init

2. Install dependencies: Ensure you have the required Python packages installed. You can install them using pip:

   python -m pip install -r requirements.txt

3. Configure Dataset Paths: Open config.py and update the paths to point to your local copy of the PAD-UFES-20+ dataset.

   # config.py
   from pathlib import Path
   
   DATA_PATH = Path('data')
   PAD_20_PLUS_PATH = Path("/path/to/your/pad-ufes-20-plus") # Update this line

To run the k-fold cross-validation experiments, use the benchmarks/kfold.py script. The validation metrics are saved in benchmarks/pad20plus/results/.

python -m benchmarks.kfold

The project currently supports the model MobileNet-V3 (defined in models/models_hub.py). It also supports feature fusion using MetaBlock to combine image features with clinical metadata.