The glottal midline and the glottal area are two important features for analyzing the voice. Here we provide GlottisNetV2, an open-source deep learning architecture to estimate the glottal midline and the glottal area.
GlottisNetV2 can be installed directly from the GitHub Repository.
pip install git+https://github.com/ankilab/GlottisNetV2.git
Due to version conflicts inside the pip install environment, please install the following packages according to your system/needs:
- TensorFlow (with or without GPU support) in v.2.5+
- segmentation_models
pip install segmentation-models - segmentation_models_3D
pip install segmentation-models-3D
You find all the documentation here: https://glottisnetv2.readthedocs.io/en/latest/