Future Access Enablers for Ubiquitous and Intelligent Infrastructures. Third International Conference, FABULOUS 2017, Bucharest, Romania, October 12-14, 2017, Proceedings

Research Article

Neural Network Based Architecture for Fatigue Detection Based on the Facial Action Coding System

Download
267 downloads
  • @INPROCEEDINGS{10.1007/978-3-319-92213-3_18,
        author={Mihai Gavrilescu and Nicolae Vizireanu},
        title={Neural Network Based Architecture for Fatigue Detection Based on the Facial Action Coding System},
        proceedings={Future Access Enablers for Ubiquitous and Intelligent Infrastructures. Third International Conference, FABULOUS 2017, Bucharest, Romania, October 12-14, 2017, Proceedings},
        proceedings_a={FABULOUS},
        year={2018},
        month={7},
        keywords={Neural networks e-health Bioengineering Facial expression recognition Image processing},
        doi={10.1007/978-3-319-92213-3_18}
    }
    
  • Mihai Gavrilescu
    Nicolae Vizireanu
    Year: 2018
    Neural Network Based Architecture for Fatigue Detection Based on the Facial Action Coding System
    FABULOUS
    Springer
    DOI: 10.1007/978-3-319-92213-3_18
Mihai Gavrilescu1,*, Nicolae Vizireanu1,*
  • 1: University “Politehnica” of Bucharest
*Contact email: mike.gavrilescu@gmail.com, warticol@gmail.com

Abstract

We present a novel non-invasive neural network based three layered system for detecting fatigue by analyzing facial expressions evaluated using the Facial Action Coding System. We analyze 16 Action Units pertaining to eye and mouth regions of the face. We define an Action Units map containing Action Unit intensity levels for each frame in the video sequence and we analyze this map in a pattern recognition task via a feed-forward neural network. We show that emotion-induced frontal face recordings offer more information in the training stage, while for testing stage the random dataset can be used with no major impact on accuracy, specificity and sensitivity. We obtain over 88% accuracy in intra-subject tests and over 83% for inter-subject tests and we show that our system surpasses the state-of-the-art in terms of accuracy, specificity, sensitivity and response time.