This University of Melbourne research project explores how novice audio describers interact with and revise AI-generated audio descriptions (AD) for factual videos.
The study examines participants’ decision-making processes when evaluating and editing AI-generated descriptions. Participants will also compare their revisions with professionally produced descriptions, review authentic feedback from blind and visually impaired (BVI) audiences, and reflect on their experiences.
The study session, including the experiment and interview, will be conducted entirely online and will take no more than 90 minutes. Participants will receive an AUD $100 (or equivalent) voucher in appreciation of their time and contributions.
Do you have less than two years of experience in producing audio descriptions in English? Interested in participating? Please contact Dr Yu Hao at haoyh2@unimelb.edu.au for more information.