Series: The Work of the Smith-Kettlewell Institute
Part II: The Video Description Research and Development Center
Although description (sometimes called audio description, video description, or a handful of other terms) didn't become formalized until the 1970s and 1980s, Josh Miele, a research scientist and principal investigator for the Smith-Kettlewell Institute in San Francisco, surmises that it was probably in effect a few thousand years ago. If a guy who was blind was in the crowd at one of the Greek tragedies, he proposes, that guy was probably poking his buddy in the ribs, demanding to know what in the world was going on down there.
Video description acquired national acclaim with the work of WGBH Boston when professional writers, describers, voice talent, and engineers began producing formalized description for PBS television programs and, later, Hollywood movies. The purpose of video description is to fill in the visual gaps for viewers who are blind or visually impaired. The more complex television and movies become and, thus, the less likely it is that an individual can speculate what the visual components of the program might be, the more essential video description becomes if viewers who are blind are to have equal access to the same content as their sighted colleagues and friends.
Josh Miele says that his latest project, crowd-sourced description for all types of video content, may well be his most notable contribution to date.
The Prevalence of All Things Video
For the last decade, Miele, who holds a PhD in psychoacoustics, has been conducting projects at Smith-Kettlewell that employ existing technologies to solve problems he and others who are blind have encountered.
Miele says that looking at ways to make video description more readily available was a natural route for him to take, since video is currently prevalent and becoming ubiquitous. There is video in education, employment, and entertainment. While it has long played a role in popular art and culture, video today has made its way into the classroom, web training, standardized testing, and, of course, myriad social media applications that range from newsworthy to just plain fun.
In other words, the rate at which video is produced is dramatically outpacing the rate at which description for the blind consumer can be produced. The solution, as Josh Miele sees it, is to tap into that centuries-old tradition of using amateur describers, essentially crowd-sourcing description or, to put it another way, poking sighted people in their virtual ribs.
With grant funding from the US Department of Education's Office of Special Education Programs (OSEP), the Video Description Research and Development Center was formed two years ago. Josh Miele serves as its director.
Recognizing that an excellent outcome is made more probable by bringing as many interested parties to the table as possible, Miele also formed the Description Leadership Network (DLN). Now boasting 10 organizational members with 70 attendees at its most recent in-depth meeting, the DLN counts among its constituents most of the major organizations with an interest in blindness and/or video description.
What if, the researchers asked, that pool of amateur describers everywhere (the friends and family members of all those people poking their neighbors in the ribs to find out what's happening) could be harnessed to describe video that could be accessed when it was actually needed?
YouDescribe and the Descriptive Video Exchange
To date, the most visible outcome of Smith-Kettlewell and the VDRDC has been the development of the Descriptive Video Exchange (initially funded by the National Eye Institute) and YouDescribe. Miele reasoned that the constantly growing body of online video content on YouTube is the most visible and perhaps largest aggregation of video content currently in need of description. If you want to know anything from how to make a square knot to how to build a coffee table to how to concoct the perfect margarita, chances are that you can find a demonstration on YouTube. However, directions such as "wrap the string in this direction" or "steady the wood as you see me doing here" don't convey much usable information to the knowledge seeker who can't see the computer screen.
The idea with YouDescribe, is that an accompanying description for any YouTube video could be recorded by anyone and added to the Descriptive Video Exchange. A person who is blind could click on any YouTube video to find out if it has an accompanying description in the DVX. If it has two or three descriptions, such as one crafted by an architect and another by a fashion designer, the user could select the description of greatest personal appeal. If there is no description yet available, the consumer could send a link of the video and a link for YouDescribe to a friend or colleague (an amateur describer) who can describe what is being shown. That newly created description can then be added to the DVX database for access by future users.
Description Present and Description Yet to Come
What Josh Miele and Smith-Kettlewell have accomplished thus far is to build the application program interface (API) to make YouTube work with the video description component. YouDescribe has some advantages over traditional video description methodologies. For example, rather than working description into available pauses as describers of live theater and film have done, YouDescribe allows for what Miele is calling "extended description." If, in other words, the existing video has only a three second pause but adequately describing a particular action or concept will require 13 seconds, the YouTube video can be paused to allow for that extended description. Miele concedes that this extended description concept, while appealing to the individual consumer who is blind, would probably not work well when watching a video for entertainment with sighted family members or friends. For educational contexts, however, the notion of taking the time needed to describe an integral visual concept on the screen without sacrificing the original audio content has obvious appeal.
For now, a few dozen samples of YouDescribe working in conjunction with YouTube comprise the DVX. The tools aren't quite ready for genuine crowd-sourcing yet, but when they are, individuals should be able to click on a given YouTube video, record a description with YouDescribe, and have it added to the DVX database. People might even be able to create wish lists, suggesting content for those amateur describers open to the challenge of new description frontiers.
Miele is quick to point out, however, that the DVX is not only for amateurs. "It has great potential as a platform for distribution of professional description as well," he says. "DVX offers the possibility of pay-for-view description that could be the basis of a whole new revenue model for existing description professionals as well as for a growing cottage industry of part-time, at-home describers."
Smith-Kettlewell and the VDRDC are in the business of research and development, not the business of manufacturing and distributing products. Their role is to build the tools to solve a problem in the hope that others will pick up the knowledge and run with it.
At this point, YouDescribe works with one player, YouTube. Eventually, Miele envisions a time when the same approach to crowd-sourcing description could work with other players as well. Imagine, for instance, popping a movie into your DVD player and having that player check the Internet for accompanying video descriptions available in real time. Imagine downloading a favorite TV show via Netflix, ROKU, or Apple TV and simultaneously downloading the description recorded by a describer somewhere.
Josh Miele and his team at Smith-Kettlewell are developing the tools that could transform such fantasies into possibilities.
Comment on this article.
Previous Article | Next Article |
Table of Contents
Copyright © 2013 American Foundation for the Blind. All rights reserved. AccessWorld is a trademark of the American Foundation for the Blind.