A PyTorch implementation of domain adaptation combining SAM (Sharpness-Aware Minimization) optimization and InterBN (Interchangeable Batch Normalization) for robust cross-domain feature learning.

• SAM Optimization: Implementation of Sharpness-Aware Minimization with AMP support
• InterBN: Novel batch normalization approach for domain adaptation
• Mixed Precision Training: Automatic Mixed Precision (AMP) support
• Advanced Training Features:
  • Exponential Moving Average (EMA)
  • Cosine learning rate scheduling
  • Sparse regularization
  • Domain adversarial training
• Experiment Tracking: Support for TensorBoard and Weights & Biases
• Development Tools: Testing, formatting, and linting support

CFSAM_InterBN_Project/
├── configs/                    # Configuration files
│   └── default_config.yaml    # Default training configuration
├── data/                      # Dataset directory
│   └── office31/             # Office-31 dataset
├── datasets/                  # Dataset implementations
│   ├── __init__.py
│   ├── office31.py           # Office-31 dataset loader
│   └── transforms_config.py   # Data augmentation configs
├── models/                    # Model implementations
│   ├── __init__.py
│   ├── feature_extractor.py  # Feature extraction network
│   ├── classifier.py         # Classification head
│   ├── domain_discriminator.py # Domain discriminator
│   └── interbn.py           # InterBN implementation
├── utils/                    # Utility functions
│   ├── __init__.py
│   ├── losses.py            # Loss functions
│   ├── sam_optimizer.py     # SAM optimizer
│   └── ema.py              # EMA implementation
├── scripts/                 # Training scripts
│   ├── train.py            # Main training script
│   └── evaluate.py         # Evaluation script
├── logs/                   # Training logs
├── checkpoints/           # Model checkpoints
├── tests/                # Unit tests
├── requirements.txt      # Project dependencies
└── README.md            # Project documentation

1. Clone the repository:

git clone https://github.com/yourusername/CFSAM-InterBN.git
cd CFSAM-InterBN

2. Create a virtual environment (recommended):

python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install dependencies:

pip install -r requirements.txt

1. Download the Office-31 dataset
2. Extract it to data/office31/
3. Organize the data as follows:

data/office31/
├── amazon/
│   └── images/
├── dslr/
│   └── images/
└── webcam/
    └── images/

1. Basic training:

python scripts/train.py

2. Training with custom config:

python scripts/train.py --config configs/your_config.yaml

• --lr: Learning rate (default: 1e-3)
• --batch_size: Batch size (default: 32)
• --epochs: Number of epochs (default: 20)
• --rho: SAM rho parameter (default: 0.05)
• --ema_decay: EMA decay rate (default: 0.9998)

To evaluate a trained model:

python scripts/evaluate.py --checkpoint path/to/checkpoint.pth

• Format code:

black .
isort .

• Run tests:

pytest tests/

Performance on Office-31 dataset (Accuracy %):

This project is licensed under the MIT License - see the LICENSE file for details.

1. Fork the repository
2. Create your feature branch (git checkout -b feature/amazing-feature)
3. Commit your changes (git commit -m 'Add some amazing feature')
4. Push to the branch (git push origin feature/amazing-feature)
5. Open a Pull Request