Library to provide models trained on the VVAD-LRS3 Dataset. The library also contains preprocessing pipelines. Applications are Speaker detection in scenarios, where multiple people are in the robot's field of view and stare detection for proactive approaches.

vvadlrs3 depends on dlib which needs build tools to be installed over pip. Here is described what is needed.

On Ubuntu, you need to install the following:

sudo apt-get install build-essential cmake libopenblas-dev liblapack-dev libx11-dev libgtk-3-dev

pip install vvadlrs3

The models are trained on the VVAD-LRS3 Dataset

^{Some samples visualized. Samples with green borders are positive samples, samples with red borders are negative samples}