CaptionMeNow

The Web began as a visual medium that has expanded in recent years to include rich media content such as audio and video. While this transformation has added value for many, it has proved problematic for other users who are deaf, hard of hearing or simply prefer to process information visually. Now with the touch of a ‘CaptionMeNow’ button, Web users can obtain a written transcript of audio on demand. IBM CaptionMeNow uses speech-to-text technology to generate written content quickly and more cost effectively than before. CaptionMeNow also automatically synchronizes the transcript with audio and reintegrates it into the appropriate multimedia format.

For users who are deaf or hard of hearing, the benefits of CaptionMeNow are clear. The ability to obtain a written transcript of audio and video content provides them with a critical tool to enhance Web usability. However, research indicates that for many people reading text is faster than listening to an auditory equivalent. So CaptionMeNow can help improve efficiency and comprehension for a much broader Web audience that includes hearing people, second language learners and people with learning disabilities who understand and prefer written language to spoken language.

CaptionMeNow’s written transcripts offer advantages beyond enabling rich-media access for more users. With a written transcript, users can more easily search audio content for areas of particular interest. And since text-based search is more expedient than audio-based search, CaptionMeNow can help you and visitors to your Web site quickly archive and retrieve audio content stored in large media libraries. Finally, transcribed text can be sent more easily to automatic text summarization or translation programs, helping you reach out to global markets.

In the past, captioning for audio and video content was a costly, labor intensive task. CaptionMeNow offers semi-automated captioning that is supplemented with self-editing tools and shadowing when necessary. So the entire process requires less human intervention than before, making it an affordable option for more businesses and educational institutions. For example, when the speech quality of an audio or video clip is adequate, captioning occurs automatically. Real-time editing capabilities can be added to help ensure the highest quality transcript. When audio speech is unclear, the content is evaluated by automated selection criteria that routes it to the most efficient, cost-effective transcription means available.

For more information on CaptionMeNow and other accessibility research technologies, contact Sara Basson at sbasson@us.ibm.com