Stanford Captioning System:
A Workflow Model
for Producing Captioned Media
Brief History
- John Foliot approached the Office of Accessible Education with an idea to streamline the production of captioned videos at Stanford University
- The proposal aimed to address institutional challenges and simplify technical issues related to captioning
- Funding was identified through the Center for Universal Design in Education [OAE] in Fall Quarter, 2008
- Development of the Stanford Captioning System began in Winter Quarter, 2009
Emphasis was placed on developing of a workflow solution to simplify the process of creating accessible media as well as promote captioning benefits to content producers.
Institutional Challenges
Preconceptions and Assumptions
- captions are costly to produce
- requires technical expertise and knowledge of captioning applications
- involves large amounts of time to produce captioned media
- perceived as not providing much benefit outside of "accessibility"
Conceptual Overview
![[Workflow path of the Stanford Captioning System]](images/captionsystemfinal.png)
Using the Captioning System
Starting a New Project
![[Screenshot of new project input fields]](images/upload1.png)
The system currently accepts the following formats:
- MP3, WMA, WAV
- WMV, MOV, MP4
File sizes are currently limited to 1GB in size.
For media files that are greater than 1GB, uploading just the audio format is an option.
Choosing a Delivery Priority
![[Screenshot of the priority selections for transcription services]](images/upload2.png)
Content authors can choose the "turnaround time" for their project.
The system currently includes options for:
- 1-Business Day
- 2-Business Days
- 5-Business Days
The cost per minute is less for longer "turnaround times".
Selecting a Vocabulary
![[Screenshot of the vocabulary settings with special terms input field]](images/upload3.png)
Vocabulary terms are included in the e-mail to the transcription company.
Additional vocabulary terms may also be included in the message and are appended to an online text-file.
Results
Video Processing
After the video file has been uploaded and processed, the Captioning System will display the results of the project as the individual steps are completed.
Transcript Processing
![[Screenshot of the Reults page showing Transcription and Caption Status]](images/results2.png)
When the time-stamping has been completed, the files are available in several formats.
An e-mail notification is also sent to the content author with links to the media and caption files.
Playback on Web Pages
HTML Code
![[Screenshot of the HTML code to embed into a Web page]](images/results3.png)
Web Page Code is for use on basic Web pages, Dreamweaver templates, etc.
Embed Code for Blogs is for use in content management systems (e.g., Drupal), blogs, or on social networking sites.
Delivery Support
Customized Media Player
A customized version of the JW FLV Player is provided to ensure an accessible media player for the Web page delivery.
Current Examples
Haas Center
Graduate School of Business
Stanford University YouTube Channel
Future Considerations
Captioning System
- Allow for "bypass" option for video encoding and additional formats
- Providing different video resolutions (4:3 vs. 16:9)
- Improve synchronization between captions & video
HTML 5 <video>
- Provide copy/paste code that includes HTML5
- Support Ogg/Theora format exports
Related Investigations
Video Search
- Google
- Yahoo!
- ReelSurfer
- Customized Implementations
Thank You
Additional Web Resources
Stanford Captioning System![[Submit the Content]](images/submit.png)