/Face-img-video-based-drowsiness-recognition

This repository is dedicated to my collaboration in the "AUTOMOTIVE" Project. This project's objective is to development automatic face image/video-based drowsiness recognition.

Primary LanguageJupyter NotebookGNU General Public License v3.0GPL-3.0

Face Img/Video based drowsiness recognition

This repository is dedicated to “AUTOMOTIVE” Project. This was a collaboration with Visual Computing and Machine Intelligence (VCMI) Research Group of | INESC Technology and Science | Porto.

Goal

Development of robust real-time Drowsiness detector system to detect driver fatigue:

  • Explore impact of facial dynamic fusion information (landmarks and eyes/mouth images)
  • Compare the use of DBN with traditional ML classifier (SVM, KNN, DecisionTree)

image

Approach

Development ML\DL models for detecting drowsiness recognition based on facial image/video.

Scripts/Files

  • pp_presentation (pdf/slides that summarize the internship results and were used to support the oral presentation)
  • python_notebook (python implementation: preprocessing, frames visualization, Training loop and testing)