Scheduled at 9:45 in Matchless on Wednesday, November 18.
#43273Speaker(s)
- Joel Snyder, PhD, President - Founding Director Emeritus, Audio Description Associates, LLC - Audio Description Project of the American Council of the Blind
Session Details
- Length of Session: 1-hr
- Format: Lecture
- Expertise Level: Beginner
- Type of session: General Conference
Summary
A review of audio description and the use of AI: AI apps are inevitable and will surely bring great advances to humanity. But—at least for the foreseeable future—is experienced and human talent key to effective audio description (AD)? Can AI contribute to the writing and speech/oral interpretation techniques/nuances provided by human analysis and production?
Abstract
It seems as though we are on the cusp of AI dominance in all manner of endeavor. Speech synthesis is already employed by some companies who produce AD for broadcast television. But the writing of audio description is dependent on an understanding and thorough analysis of the work to be described. Similarly, the appropriate voicing of AD is done with nuance, attention to the images on screen, and an understanding of the phrasing used in the writing of the description. How can audio describers work with/benefit from the incorporation of AI in the AD production process? Can the spread of AI/speech synthesis be stopped or forestalled? Should it be?
It is notable that in 2021, the American Council of the Blind passed a resolution noting its “full support for … the use of human voices in the voicing of audio description for cinema and narrative video or streaming.” https://www.acb.org/2021-resolutions#Resolution2122
This paper/presentation will offer AI inquiries posed to ChatGPT and Eleven Labs regarding the effectiveness of AI’s drafting and voicing of AD as well as a comparison of human and AI-generated AD excerpts—its writing and voicing.
The session will draw from the speaker's 45 years of experience in the production of audio description and the training of describers throughout the United States and in over 60 countries. The focus of the session will be on how AI can be used as a partner in the creation of AD and thus speed its development while always incorporating established AD fundamentals and ensuring high-quality AD. AI apps are inevitable and will surely bring great advances to humanity. But—at least for the foreseeable future—experienced and *human* writers and voice talents are key to effective AD.
Keypoints
- Writing of AD is dependent on an understanding and thorough analysis of the work to be described.
- The voicing of AD should be done with nuance, attention to images on screen, and an understanding of phrasing.
- How AI can be used as a partner in the creation of AD.
Disability Areas
Vision
Topic Areas
Artificial Intelligence, Assistive Technology, Uncategorized
Speaker Bio(s)
Joel Snyder, PhD
Dr. Joel Snyder is known internationally as one of the world’s first “audio describers,” a pioneer in the field of Audio Description, a translation of visual images to vivid language for the benefit, primarily, of people who are blind or ahve a vision impairment. The visual is made verbal—and aural, and oral. Since 1981, he has introduced audio description techniques in over 40 states and 63 countries and has made hundreds of live events accesible. In 2014, the American Council of the Blind published Dr. Snyder’s book, The Visual Made Verbal – A Comprehensive Training Manual and Guide to the History and Applications of Audio Description, now available from the Library of Congress as an audio book voiced by Dr. Snyder and in Braille, in screen reader accessible formats, and in English, Polish, Russian and Portuguese (Spanish, Greek and Chinese are in development). Dr. Snyder is the President of Audio Description Associates, LLC and he serves as the Director of the Audio Description Project of the American Council of the Blind (www.acb.org/adp).