SimpleDet is a custom object-detection toolkit built around MMDetection/OpenMMLab workflows for training, inference, evaluation, and experiment packaging.

Base package only:

python -m pip install simpledet

Supported CPU runtime for the public detection APIs:

python -m pip install "simpledet[cpu]"

Backward-compatible alias:

python -m pip install "simpledet[openmmlab]"

Optional workflow extras:

python -m pip install "simpledet[geo,plots]"

From source while iterating locally:

git clone https://github.com/sirbastiano/MDet.git
cd MDet
python -m pip install -e ".[cpu]"

Sanity checks:

python -m simpledet --version
python -m simpledet --check-openmmlab

The package is published as a pure-Python wheel. The supported runtime contract is:

• Linux, macOS, and Windows
• Python 3.10, 3.11, and 3.12
• CPU-only dependency stack for simpledet[cpu]
• wheel-only dependency resolution for release verification

If a dependency cannot be installed from wheels on a claimed platform/Python pair, that matrix entry is not considered supported.

make venv
make sync
make sync-cpu
make build
make check

• simpledet/: installable package
• simpledet/src/: packaged configs and custom model components
• tests/: unit and packaging checks
• docs/: project documentation site sources

Local release verification:

python3 -m build
python3 -m twine check dist/*
PYTHONPATH=simpledet python3 -m unittest discover -s tests -p 'test*.py'

The CI release workflow builds the sdist/wheel, runs the test suite, and verifies wheel-only installation of the built artifacts on Linux, macOS, and Windows for Python 3.10-3.12.

This project is licensed under the MIT License.