Accessible lectures: moving towards automatic speech recognition models based on human methods

Miltiades Papadopoulos, Elaine Pearson

    Research output: Contribution to conferencePaperpeer-review

    71 Downloads (Pure)

    Abstract

    The traditional lecture remains the most common method of teaching and while it is the most convenient from a delivery point of view, it is the least flexible and accessible. This paper responds to the challenge of meeting the needs and access requirements of students with disabilities by urging further adaptations in the learning environment. The aim of this work is to explore the way speech recognition technology can be employed in the University classroom to make lectures more flexible and accessible. The concluding section explores the concept of an ASR model, based on principles derived from studies of human methods of recognition, in order to increase their performance and efficiency.
    TYPE:
    Original languageEnglish
    Pages273
    DOIs
    Publication statusPublished - 2008
    EventThe 10th international ACM SIGACCESS conference - Halifax, Nova Scotia, Canada
    Duration: 13 Oct 200815 Oct 2008
    Conference number: 10

    Conference

    ConferenceThe 10th international ACM SIGACCESS conference
    CountryCanada
    CityNova Scotia
    Period13/10/0815/10/08

    Fingerprint Dive into the research topics of 'Accessible lectures: moving towards automatic speech recognition models based on human methods'. Together they form a unique fingerprint.

    Cite this