Voice to Text

Posts about deaf-friendly voice-to-text technology.

When Google first announced Project Glass, a prototype augmented reality device that appears to be the company’s first foray into wearable computing, I immediately saw the potential for this as an accessibility device.  I’m sure Google did as well, as the patent they filed seemed to hint at it being used as a speech-to-text display and visual alert for the deaf.

Some visualizations I came up with for the potential of Project Glass as a deaf-tech device include:

  • Ability to download and load subtitle files in the glasses at movies, plays, and other scripted performances
  • Speech-to-text “virtual live captions” at lectures and in group conversations
  • Visual alerts for audible alerts or dangers

{ 0 comments }

Although I can lipread quite well and can successfully hold a conversation in most one-on-one situations, following groups of people talking is very difficult.  Trying to focus on a single person giving a long lecture is also difficult, which is why I had notetakers in all my classes in school, and if I go to a conference, I always request a CART captioner.  It’s long been a dream of mine to have a handheld device that could pick up spoken text around me and translate it to captions, so I could never be left out of conversations again.

We haven’t quite gotten there yet – maybe someday – but speech recognition technology has been improving by leaps and bounds.  CPC has introduced a new live captioning via speech recognition product called YouCaption, which aims to  be a low-cost and easy to use real-time captioning software program.  The software itself is Windows-only, but the results can be displayed on a website, computer screen, smartphone, and tablet.  From the press release:

Captions PowerPoint®, Webinars, On-Line Meetings and More
Captions are displayed on PowerPoint® Presentations and computer screens. If you use online meetings or a screen sharing system such as WebEx®, GoToMeeting®, TeamViewer®, etc., you can include the closed captions window as part of your shared screen. Traditional hardware encoder based systems cannot easily display captions to these popular platforms.

Captions Web Sites & Mobile Devices
The captions can also be directed to a live web page which can be viewed on virtually all web browsers, including smart phones and tablets like the iPad® and iPod®, with no special software or plug-ins needed. This is ideal for reaching a mobile or remote audience, especially when bandwidth may be an issue. (This feature requires access to a web server with PHP and MySQL installed.)

Faster Real-time Closed Captions
Captions are displayed moments after the corresponding words are spoken, just like they are with steno based captioning systems. There is no buffering or need to pause, which are pitfalls of many previous speech recognition based systems.

Live TV Broadcast
YouCaption can interface with CPC’s award winning CaptionMaker software to control a local or remote broadcast closed caption encoder.

Post Production
YouCaption outputs a timed transcript of your session. If you need to broadcast or upload a video recording of your event later, you can use this script in CPC’s CaptionMaker® or MacCaption™ closed captioning software to easily add captions to the recording. You can also provide a plain text transcript to anyone who missed your event, or use it for indexing and archiving.

 

{ 0 comments }

GotVoice – read your voicemail

by Kathryn Hill on October 4, 2007

GotVoice

GotVoice is a service that converts voicemails to text and sends them to you via email or text messages. It’s a paid service, but very useful if you’re deaf and need to be in contact with hearing people who don’t do Relay services, email or text messages.

{ 1 comment }

Text transcripts on LiveJournal voice posts

by Kathryn Hill on May 23, 2007

The LiveJournal blogging platform offers members the ability to do voice posts, and text transcripts are either made by the creator of the podcast or the readers; it’s a community-driven effort and works quite well – many people are willing to do this. Podcasts that are under three minutes are automatically transcribed by SpinVox.

Previously:
Spinvox: voicemail to text service
Casting Words transcripting services
Project ReadOn – free web captions
Speche Communications: real time text streaming
Jott’s potential to transcribe podcasts is unrealized
Earize Text Streaming for Internet Radio
Podcast transcripts for the deaf

{ 0 comments }

Voicewriting

by Kathryn Hill on April 19, 2007

Voicewriting is an emerging interpreting technology. A transcriptionist uses a stenomask to repeat spoken words without being heard by others. The stenomask is connected to a laptop computer which has speech recognition software installed in it, and a text transcript of the conversation is automatically created. Since the stenomask allows full focus of only the transcriptionist’s voice, the speech recognition software program is able to provide more accurate results.

{ 0 comments }

Jott, which I blogged about earlier today, has another feature where users can leave a short voice message that will be typed up and posted to their blog.  It only allows about 30 seconds of spoken text, which is a shame; if the time was unlimited, this would solve the issue of a lack of text transcripts available for podcasts.

{ 0 comments }

Text messaging is a wonderful tool for the deaf and hearing impaired. I was one of the first people among my peers to have a cell phone that had SMS/text messaging. Since I’m text-heavy, I made sure to invest in a phone that had a QWERTY keyboard. I excitedly sent text messages to all my friends, only to have them complain that sending a reply on their slower numeric keypads was too much trouble for them. The texting fad hadn’t caught on yet at that point.

Nowadays, almost everyone I know has embraced text messaging, apart from a select few who are either using antiquated cell phones or have cell carriers that charge extra for text messages.

That’s where Jott comes in – among several features, it also offers “hands-free messaging.” Jott is very useful for people who can’t send text messages on their phone for whatever reason.

How it works: sign up for a Jott account (it’s free) and add the Jott number to the speed dial on your phone (877 568 8486.) Next, import your contacts into your Jott profile. Each contact is assigned a name (“Mom,” “Cthulhu,” etc.) Groups can even be assigned. To send a text message, simply dial the number, state the name of the contact or group you wish to address, and leave a voice message. Your message will be typed up and SMS’d/texted to your recipitent. Easy!

Now hearing people who previously could not sty in touch with their deaf friends or family members because they couldn’t send SMS/text can now do so.

{ 4 comments }

SimulScribe – voice mail transcribing service

by Kathryn Hill on April 3, 2007

SimulScribe is a voice mail transcribing service that is primarily aimed at hearing people who are too busy to listen to their voice mails. For a small fee, voice mails are transcribed into text and either emailed to the subscriber, or sent in text message format. These services are provided free already by IP-Relay and i711, so I am not convinced the deaf market would be interested in this service, but you never know.

EDIT 04/05/07: It occurred to me that this would be an useful service for deafies who have Skype.

{ 1 comment }

Spinvox: voicemail to text service

by Kathryn Hill on March 15, 2007

Spinvox is a new service that provides speech-to-text service for voice mails. Voice mails are typed up and emailed or text messaged to subscribers for a fee. While deafies get this service free already with IP-Relay and i711, what makes Spinvox stand out is that they’re working on Skype integration.

{ 0 comments }

Earize Text Streaming for Internet Radio

by Kathryn Hill on January 23, 2007

I just discovered a new technology called Earize, which calls itself an internet text streaming service. From their website:

Earize is an Internet streaming text service. This service can be used exclusively for caption radio, and to assist deaf and hard of hearing people in meetings, telephone conferences, and seminars, to name a few. Audiences can be increased by 8.6% just by reaching the folks with hearing loss. The service is also a vehicle for additional advertising to this sometimes overlooked segment.

Users can read the text from a computer, PDA, or cell phone, and can customize the text by font size and color. Also, if something is missed, users can scroll back and read the transcript. Broadcasters pay a small fee to have their show transcripted; the service is free for the receiver.

{ 0 comments }