German Sign Language Detector

This is the supplementary proof of concept to show that the thesis which was claimed in the bachelorthesis is correct.

Installation

If you want to run this program on your own machine use the package manager pip to install the dependencies:

git clone https://github.com/Birkenpapier/GSL-Detector.git
cd GSL-Detector
pip3 install -r requirements.txt

Usage

There are different ways to utilize the work in this repository

Translate German Sign Language

To use this program you can use the pretrained weights action.h5 in this repository.

python3 3_Real_Time.py

Training

Either you'll use the precaptured training data in MP_Data to evaluate the results:

python3 2_Train_Model.py

or you can capture new data for the hardcoded sign language actions ['hallo', 'danke', 'vielglück', 'bitte', 'wo']:

python3 1_Get_Data.py

Docker

If decide to take the easiest way to setup the project feel free to use the provided dockerfile and startup shell script in the repository:

sudo docker build -t gsl-detector . && sh runDocker.sh

Please make sure that the video device is attached and accessible from /dev/video0! After the successful start you can use the detector to detect one of the predefined actions:

Contributing

Pull requests are welcome. For major changes, please open an issue first to discuss what you would like to change.

License

unlicense

Citation

  @misc{birkenpapier2022gsldetector,
    author = {Kevin Peivareh},
    title = {German Sign Language Detector},
    year = {2022},
    howpublished = {\url{https://github.com/Birkenpapier/GSL-Detector}},
    note = {commit xxxxxxx}
  }

Attribution/Thanks

• This project could not have happened without the tutorial and explanations given by Nicholas Renotte.