Contributing

July 11, 2026 · View on GitHub

Thank you for your interest in contributing to OrientedDet!

Getting Started

  1. Fork the repository
  2. Clone your fork: 
    git clone https://github.com/DL4EO/oriented-det.git
cd oriented-det
  3. Create a venv and install in development mode (uv): 
    uv venv --python 3.12
source .venv/bin/activate
uv pip install -r requirements.txt
uv pip install -e ".[dev]"

Development Setup

Running Tests

CI runs pytest tests/ on GitHub Actions (CPU PyTorch). Locally:

# Run all tests
pytest

# Run specific test file
pytest tests/test_geometry.py

# With coverage
pytest --cov=oriented_det tests/

Code Style

We use black for formatting and ruff for linting:

# Format code
black oriented_det/ tests/

# Lint
ruff check oriented_det/ tests/

Contribution Guidelines

Code Style

  • Follow PEP 8
  • Use type hints where appropriate
  • Write docstrings (Google style)
  • Keep functions focused and small

Testing

  • Add tests for new features
  • Ensure all tests pass
  • Aim for good test coverage

Documentation

  • Update docstrings for new functions/classes
  • Add tool scripts or examples to user guide if applicable
  • Update API reference

Pull Requests

  1. Create a feature branch
  2. Make your changes
  3. Add tests
  4. Update documentation
  5. Ensure all tests pass
  6. Submit a pull request

Project Structure

oriented_det/
├── geometry/      # Core geometry primitives
├── ops/           # IoU and NMS operations
├── data/          # Dataset loading and utilities
├── models/        # Model architectures
├── train/         # Training utilities
└── utils/         # Configuration and visualization

Areas for Contribution

Aligned with the Roadmap:

  • v0.2–v0.3: Faster R-CNN probiou Hub publish; Rotated FCOS model and configs
  • v0.4: HRSC2016 and FAIR1M dataset loaders
  • v0.5: RTMDet-R and native YOLO-OBB (clean-room, no Ultralytics)
  • Docs: Tutorials, hosted MkDocs, eval reports
  • Ops: Fused CUDA rotated IoU/NMS kernels
  • Bug fixes and tool scripts

Questions?

Feel free to open an issue for questions or discussions!