Automatic vs. Human Captioning

Automatic captioning

Fully automatic speech recognition technology creates captions in real-time. Automatic captioning can be useful to most people, including hearing individuals who benefit from audio prompts, such as some people with learning disabilities and some people who speak English as an additional language. It can also be useful in situations where audio quality is not optimal. The accuracy of automatic captioning varies widely based on speaker, mic-placement, accent, terminology, and other uncontrollable factors.

Human captioning

Human captioning involves a human typing in real-time while someone is speaking or while a video is playing. Human captioning is essential for Deaf and some Hard-of-Hearing people to fully participate.

Live Captioning Guides

Captioning Type by Audience

Audience Zoom:
Automatic Captioning
 		Google Meet:
Automatic Captioning
 		Google Slides:
Automatic Captioning
 		Zoom: 
Human Captioning
Can be used in classroom Yes,
unless there is a student with a captioning accommodation
 Yes,
unless there is a student with a captioning accommodation
 Yes,
unless there is a student with a captioning accommodation
 Yes
Can be used in small groups
 Yes,
if no one requests human captioning
 Yes,
if no one requests human captioning
 Yes,
if no one requests human captioning 		Yes
Can be used within a School/Unit/Department
 Yes,
if no one requests human captioning
 Yes,
if no one requests human captioning
 Yes,
if no one requests human captioning
 Yes
Can be used for public events

Yes, if all are true:

  • RSVP required to attend
  • Accommodation statement included in promotional materials and RSVP
  • No one requests human captioning

Yes, if all are true:

  • RSVP required to attend
  • Accommodation statement included in promotional materials and RSVP
  • No one requests human captioning

Yes, if all are true:

  • RSVP required to attend
  • Accommodation statement included in promotional materials and RSVP
  • No one requests human captioning
 Yes

Captioning Features by Platform

Function Zoom:
Automatic Captioning 		Google Meet:
Automatic Captioning
  		Google Slides:
Automatic Captioning
 		Zoom:
Human Captioning
Can provide a transcript
 Yes No No Yes
Captions can be viewed in post-event recording
 Yes
(the recording is retranscribed if you have enabled audio transcription)
 N/A N/A Yes
(and the recording is also automatically transcribed if you have enabled audio transcription)
Can be corrected in real-time
 No No No Yes
Can indicate a change in speaker Yes, in the transcript window Yes, automatically labels speakers by name
 No Yes, provided by human captioning vendor
Supported in breakout rooms No N/A
NYU Google Meet does not currently support breakout rooms
 N/A Yes
Can be used outside of video conferencing
 No No
 Yes, only speaker hosting presentation No
Participants can turn captions on/off Yes Yes No Yes
Currently supports languages other than English No No No Yes, if the language is supported by the vendor
Censors content deemed to be explicit
 No Yes, uses asterisks Yes, uses asterisks
 No
Engine type (how are the captions generated)
 ASR
(Automatic Speech Recognition)
 ASR
(Automatic Speech Recognition) 		ASR
(Automatic Speech Recognition) 		Human captioner
Notes
  • Free for NYU Community.
  • Based on the Otter.ai speech recognition technology.

 
  • Free for NYU Community.
  • Max attendees: 250 people.
  • Free for NYU Community.
  • Requires Chrome browser.
 NYU-approved human captioning vendors