We are happy to receive contributions from the community. If you want to contribute, please read our contribution guidelines first (this file :^) )
As JeFaPaTo is a research project, we are not able to pay you for your contributions. However, we will mention you in the acknowledgements and you will contribute to the scientific community and many researchers.
One unique feature of JeFaPaTo is that it is split into two main parts. The first being the coding library, which is the main area of functionality and the second being the GUI, which is the main area of user interaction. This allows us to have a clear separation of concerns. If you want to contribute a new feature, please consider which part of the project it belongs to. If you are unsure, please open an issue and we will help you.
We use pytest and pytest-cov for our tests. We run the following command manually to run the tests:
pytest -rA -vv -W ignore::DeprecationWarning --cov=jefapato --cov-report term-missingThe automatic CI/CD pipeline currently only runs the tests for parts of the coding library, as in GitHub actions no OpenGL context can be created. Therefore, the extraction of the facial features and the blink detection are not tested automatically. We are working on a solution for this problem. The CI/CD tests runs the following subset of the tests:
pytest -rA -vv -W ignore::DeprecationWarning --cov=jefapato --cov-report term-missing tests/test_blinking.py tests/test_earfeature.pyTherefore, we would be very happy if you could run the tests locally before you create a pull request. If you are unsure, please open an issue and we will help you. We expect to see a coverage of a high in the coding library. If you want to contribute a new feature, please also add tests for your code and if necessary test files. A proof via an image and correct CI/CD tests is enough until we have a solution for the OpenGL context problem.
Here is the current test coverage of the coding library (as of 2024-01-22):
The coding library is the main area of functionality. It contains all the code that is needed to extract the facial features and blink detection. This part should be clearly separated from the GUI and should not contain any GUI code. Basically, this part should be usable without the GUI.
Each feature should be implemented in a separate file. For example, the blink detection is implemented in the file blink_detection.py. If you want to contribute a new feature, please create a new file and implement the feature there. If you want to contribute to an existing feature, please open an issue and we will help you.
Also your implementation should be well documented and tested. We use the Google Style Guide for our code. And please also add tests for your code and if necessary test files. We use pytest for our tests. We hope to achieve a coverage of 100% in the coding library, and are happy to receive your help.
The GUI is the main area of user interaction. It contains all the code that is needed to interact with the user. This part should be clearly separated from the coding library and should not contain any code that is not related to the GUI.
Here we are happy to receive contributions to improve the user experience. If you have any ideas, please open an issue and we will help you.
It is important to note that this part interacts with the coding library. We include a lot of user sanity input checks in the GUI, but we also need to include them in the coding library. If you are unsure, please open an issue and we will help you.
If you implemented a new feature in the coding library, but have no experience with GUIs, please open an issue and we will help you to bring your feature into the GUI :).
We use PyQt6 for our GUI to ensure cross-platform compatibility and pyqtgraph for our plots. The GUI is build up from several custom widgets to ensure reusability. If you want to contribute a new widget, please create a new file and implement the widget there. If you want to contribute to an existing widget, please open an issue and we will help you.
The GUI is currently not automatically tested. Hence, we still unfortunately have some bugs and rely on feedback of our users. If you find a bug, please open an issue and we will investigate it.
- Oliver Mothes: Initial draft and concept of the project, also the name giver of the project
- Lukas Schuhmann: Medical partner and feedback provider for the windows version
- Elisa Furche: Medical partner and feedback provider for the Mac v10 version
- Elisabeth Hentschel: Medical partner and feedback provider for the Mac v13 version
- Yuxuan Xie: Investigation of EAR score and blink detection