Captions: Computer-Generated (ASR)

doctor speaking to patient while holding a tablet to translate speech to text.

Graphic credit: Ava captioning

Screenshot of iPhone with built-in captions. Wording on screen "This is an example of apples, live, captioning. It is converting speech to text. Listening..." (note: has errors)
Apple Live Captions on iPhone
Screenshot of Google's Live Transcribe with built-in captions. Wording on screen "This is an example of Google's live transcribe. It is converting speech to text."
Google Live Transcribe on tablet

Computer-Generated Captions

(Institution provided or consumer provided). See handout below.

Captions are invaluable for people with hearing loss. Captions provide visual accommodations to supplement the spoken word.

Computerized captions are better for shorter interactions, where accuracy isn’t as important – such as patient check-ins, some meetings, or in emergency situations when CART is not available, etc.

Computer-generated captions are also called “automatic captions,” “AI captions,” “speech-to-text apps,” or ASR (Automatic Speech Recognition). Smartphones, devices, and computers have apps to convert speech to text. Accuracy depends on keeping the microphone close to each speaker, how clearly people speak, background noises, and people speaking over one another.

Captions for Presentations

Captioning Apps and Built-in Capabilities

ADA & Requesting Accommodations

Computer-generated captions can meet the requirements for “auxiliary aids and services” under Effective Communication with the Americans with Disabilities Act (ADA).

Consumers with hearing loss can be hesitant to ask for accommodations, and these accommodations are rarely offered without requests. It is imperative that staff realize the high prevalence of hearing loss in all populations and proactively offer services that enable people with hearing loss to understand the information.

Healthcare

Telehealth

  • Both CART and speech-to-text captions can be integrated into telehealth platforms, or can be shown on a second device (more complex for the patient)
  • Telehealth platforms are required to have captions by the end of 2026, under the “interoperable video conferencing” section in the FCC.

HIPAA-compliant, speech-to-text. Links go to webpages that explain the company’s privacy. These companies are willing to sign a BAA (Business Associate Agreement) with a healthcare clinic subject to HIPAA, or because of their data privacy/data design, there is no need.

  • Ava captioning. Desktop and app solutions for in-person or online needs.
  • Caption Companion. A self-contained single-purpose standalone hardware device that works completely offline without internet. 
  • TranslateLive. Real-Time Text Transcription. Conversations are transcribed live, allowing individuals with hearing loss to access to conversations and information in real time.
  • Windows 11. For Windows and Mac, is a good option if the telehealth platform doesn’t have built-in captions. The transcription is done locally on the device, meaning audio isn’t sent to the cloud (unless configured otherwise), and thus meets HIPAA compliance.

Handouts

pdf
340 KB
Adding automatic captions to presentations can benefit a wide range …
pdf
361 KB
(4 pages) Compares the differences between human-generated and computer-generated captions, …
pdf
587 KB
Frazier, et al. January 8, 2024. Committee for Communication Access …
pdf
343 KB
(2 pages) Overview information on assistive listening systems and captions. …
pdf
336 KB
(2-pages) Patients with hearing loss can be hesitant to ask …

Articles

Resources

Virtual Conferencing Resources

Professional Organizations and Resources

Please email the Center for Hearing Access with photos, videos, corrections, ideas, or additional products. Contact us (this website)

The Center for Hearing Access provides educational and informational resources and does not endorse any product, business, or service.