View Item 
        •   Utrecht University Student Theses Repository Home
        • UU Theses Repository
        • Theses
        • View Item
        •   Utrecht University Student Theses Repository Home
        • UU Theses Repository
        • Theses
        • View Item
        JavaScript is disabled for your browser. Some features of this site may not work without it.

        Browse

        All of UU Student Theses RepositoryBy Issue DateAuthorsTitlesSubjectsThis CollectionBy Issue DateAuthorsTitlesSubjects

        Can we predict susceptibility to audio signals in autonomous driving by using EEG and HMM?

        Thumbnail
        View/Open
        Thesis.pdf (1.042Mb)
        Publication date
        2018
        Author
        Tompoidi, A.
        Metadata
        Show full item record
        Summary
        The aim of this study is to investigate whether susceptibility to audio stimuli can be predicted during autonomous driving. Previous studies have shown that the susceptibility to unexpected audio stimuli is intrinsically related to attentional resources being allocated during driving. Namely, the more attentiveness is required during driving, the less susceptible a driver is to unexpected audio stimuli. In our study, we trained Hidden Markov Models in an attempt to find hidden state(s), which could be indicative of the susceptibility to potential audio signals played in autonomous cars. Retrieved transitional probabilities of hidden states showed that transitions between states are very unlikely and the driver tends to remain at the same state for some time. On the other hand, by considering only 100ms before stimulus onset of data, could not provide us significant information in regards to attentiveness or expected susceptibility to the audio stimuli, as the most frequent state’s mean value was around zero. Additionally, by comparing before and after stimulus onset states, no significant results could be retrieved. Finally, we discuss multiple reasons which might have contributed to the inability of identifying the potential information in EEG signal using HMMs.
        URI
        https://studenttheses.uu.nl/handle/20.500.12932/38133
        Collections
        • Theses
        Utrecht university logo