Menpo. The Menpo Project Python package for handling annotated data.
What is Menpo?
Menpo is a Menpo Project package designed from
the ground up to make importing, manipulating and
visualizing image and mesh data as simple as possible. In particular,
we focus on annotated data which is common within the fields of Machine
Learning and Computer Vision. All core types are
visualizing these landmarks is very simple. Since landmarks are first class
citizens within Menpo, it makes tasks like masking images, cropping images
inside landmarks and aligning images very simple.
Menpo were facial armours which covered all or part of the face and provided a way to secure the top-heavy kabuto (helmet). The Shinobi-no-o (chin cord) of the kabuto would be tied under the chin of the menpo. There were small hooks called ori-kugi or posts called odome located on various places to help secure the kabuto's chin cord.
--- Wikipedia, Menpo
Here in the Menpo Team, we are firm believers in making installation as simple as possible. Unfortunately, we are a complex project that relies on satisfying a number of complex 3rd party library dependencies. The default Python packing environment does not make this an easy task. Therefore, we evangelise the use of the conda ecosystem, provided by Anaconda. In order to make things as simple as possible, we suggest that you use conda too! To try and persuade you, go to the Menpo website to find installation instructions for all major platforms.
If you feel strongly about using Menpo with the most commonly used Python
package management system,
pip, then you should be able to install
Menpo as follows:
> pip install menpo
We strongly advocate the use of conda which does
not require compilation for installing Menpo or it's dependencies such as Numpy,
SciPy or Matplotlib. Installation via
conda is as simple as
> conda install -c conda-forge menpo
And has the added benefit of installing a number of commonly used scientific packages such as SciPy and Numpy as Menpo also makes use of these packages.
|CI Host||OS||Build Status|
Menpo makes extensive use of Jupyter Notebooks to explain functionality of the package. These Notebooks are hosted in the menpo/menpo-notebooks repository. We strongly suggest that after installation you:
- Download the latest version of the notebooks
- Conda install Jupyter notebook and IPython:
conda install jupyter ipython notebook
- Play around with the notebooks.
Want to get a feel for Menpo without installing anything? You can browse the notebooks straight from the menpo website.
Other Menpo projects
Menpo is designed to be a core library for implementing algorithms within the Machine Learning and Computer Vision fields. For example, we have developed a number of more specific libraries that rely on the core components of Menpo:
- menpofit: Implementations of state-of-the-art deformable modelling algorithms including Active Appearance Models, Constrained Local Models and the Supervised Descent Method.
- menpo3d: Useful tools for handling 3D mesh data including visualization and an OpenGL rasterizer. The requirements of this package are complex and really benefit from the use of conda!
- menpodetect: A package that wraps existing sources of object detection. The core project is under a BSD license, but since other projects are wrapped, they may not be compatible with this BSD license. Therefore, we urge caution be taken when interacting with this library for non-academic purposes.
See our documentation on ReadTheDocs
We use pytest for unit tests.
>> pytest .
from the top of the repository will run all of the unit tests.
Some small parts of Menpo are only available if the user has some optional dependency installed. These are:
- 3D viewing methods, only available if
menpo.feature.dsiftonly available if
- Some warping unit tests are only available if