At InnoCaption, we consider one of our main differences to other real-time captioning applications the ability to use both a live stenographer and Automated Speech Recognition (ASR). You can even switch between them in the middle of your phone call. For those not familiar with the difference and why it matters, we wanted to ensure we’re communicating the benefits!

The InnoCaption application acts as a central switchboard that routes calls to our large network of stenographers who all work from secure, remote locations (not in call centers) and also integrates our ASR technology as an additional source of captions. All of these systems work seamlessly, 24 hours a day, 7 days a week, so that users have an uninterrupted and seamless calling experience.

ASR is a computerized form of providing captions to our users. InnoCaption utilizes best-in-class automated speech-to-text engines to convert speech on your smartphone calls to verbatim captions on your smartphone – we call it the “InnoCaption ASR Calling Feature.” The InnoCaption ASR Calling Feature enables InnoCaption users who feel the technology meets their accessibility needs during a call to use it – and then in the event it stops meeting their needs, to seamlessly switch over to a CART provider. It is important to note that no call audio is saved by InnoCaption for machine learning purposes.

CART stands for Communication Access Realtime Translation. CART is a service in which a highly skilled CART provider, or a stenographer, listens to speech and instantaneously transcribes all the speech to text using a steno machine. A steno machine is a kind of shorthand typewriter which enables fast and accurate captions. While there are calls and situations where ASR will excel, and users will feel their accessibility needs are being met, there are also calls where a speaker’s unique speech pattern, or the call connection, will result in ASR captions which do not meet a user’s accessibility needs. For this reason, InnoCaption has implemented an in-call switching feature to enable each user to toggle between CART and ASR as dictated by their accessibility needs.

Here are a few examples where ASR or CART would be appropriate, in addition to situations where a user would want to switch between the two:

 

  • ASR-Only Example: A user would like to place a private call to his significant other. The significant other has a relatively clear voice, and the user is familiar with their voice so does not need to rely on the captions 100% for the call. The user prefers to have these calls without a live CART provider on the line and feels ASR meets their accessibility needs for the call.
  • CART-Only Example: A user has a teleconference call with multiple participants. The user chooses to use CART because they know that CART providers are able to distinguish between speakers and will prompt the user if multiple speakers are overlapping. When CART prompts the user there are overlapping speakers. The user is able to advocate for themselves by reminding participants to speak one at a time. The user feels that only CART meets their accessibility needs for this call.
  • CART to ASR Example: A user is on a professional call where there is highly technical language being used. While this user generally relies on CART for these calls and has never had an issue, they feel this particular CART provider is struggling with some of the terminology. The user clicks the CART button on the top of the screen to switch to ASR. The ASR captioning is providing accurate captioning of the speaker and the technical terms, so the user decides to stick with the ASR captioning.
  • ASR to CART Example: A user receives a call from family members for their birthday. This user starts the call using ASR captioning. Their mom has a voice which ASR understands well and the captions are fast and accurate. The call is then handed the phone to the users young niece who also wants to wish them a happy birthday, but the ASR engine is not understanding the young child’s voice.