Question · Accessibility and health
Do speech to text glasses work for deaf people?
Speech to text glasses can help some deaf and hard of hearing people follow conversations, but the evidence is thin. The strongest study we found, at a school for the deaf, reported a learning gain only in some contexts, and we did not find independent accuracy or delay tests of commercial caption glasses. Try a pair in the rooms where you will use them before you depend on them.
Updated October 16, 20264 min read9 cited facts, each traced to its source
Speech to text glasses can work for some deaf and hard of hearing people, but the research does not yet show how well they work for most people. Caption glasses display the words of the person you are talking to on a small screen in your line of sight, so you do not have to look away from the speaker. The benefit depends on three things: how accurate the captions are in your setting, how long they take to appear, and how easily you can read them while you follow the conversation.
The evidence so far comes from prototypes and small studies
The strongest study we found is a randomized field experiment at a school for the deaf. It reported that augmented reality (AR) captioning could enhance learning performance 1. The same paper found the benefit appeared only in some contexts: where the captions gave nuanced information, and for students with strong visual and auditory processing 2. That is a useful result from one school and one conference paper. It is not a promise that glasses will help every deaf or hard of hearing reader.
In the sources we checked for this page, we did not find an independent accuracy or delay test of a commercial caption glasses product. Vendor buyer's guides list accuracy, latency and battery life as things to compare, but they contain no independent measurements. Treat any figure a maker quotes as a claim to check, not a measurement you can rely on.
Delay matters more when you read captions at a glance
A caption helps only if it appears while the speaker is still making the point. A University of Toronto report on head-mounted captioning sets a benchmark: the lag should never exceed 5 seconds, and 80 percent of captions should appear within 3 seconds 3. The report is a student course project, so use the figures as a benchmark to test against, not as an established standard.
A 2026 survey of deaf and hard of hearing viewers points the same way. It found that caption latency significantly affects the viewer experience, especially the 7 to 12 second delays typical of live television 4. That survey was about TV rather than glasses. Still, a delay of several seconds can leave you reading a sentence after the conversation has moved on.
Glasses inherit the limits of the speech recognition behind them
Glasses show text produced by speech recognition, so they carry its weaknesses. In a 2024 study of eleven common services 5, streaming recognition, the mode used for live events, measured significantly lower in quality 6. A 2017 study of deaf and hard of hearing people's views on speech recognition in conversation notes that accuracy depends on speech fidelity, ambient noise and microphone quality 7.
Glasses also do not remove the information gap. A 2017 study reports that, even with accurate real-time text from human captioners, deaf and hard of hearing people receive 50 to 80 percent of the information, against 84 to 95 percent for hearing peers 8. Glasses may make it easier to keep eye contact, but the research so far does not show that they make a conversation easier to follow.
It depends on the setting
| Setting | What the evidence suggests |
|---|---|
| One speaker, close, in a quiet room | The best case for any caption display. Test the accuracy with the speakers you know. |
| Group conversation or a noisy place | Harder for the software and the reader. Text accuracy and lag time were the main obstacles in a 2017 study 9. |
| Lectures, medical visits or legal proceedings | Use a human captioner where a missed word matters. See CART captioning. |
The evidence points to the conditions in which the one randomized study found a gain: nuanced content, and readers with strong visual and auditory processing. Beyond that, the research has not yet shown who benefits most from caption glasses.
What to do before you buy or rely on caption glasses
- Test them in the real room. Use the seat, the lighting and the background noise you will have, with the people you talk with most.
- Measure the delay. Have someone speak at a normal pace and check whether the text keeps up.
- Check the errors that matter. A wrong name, number or "not" can change the meaning. Look for those as well as ordinary word errors.
- Ask the maker what was measured. Find out whether the figure covers live or recorded audio, and in which languages. For how to read these claims, see whether AI captioning is accurate enough.
- Keep a backup for high-stakes moments. Book a human captioner for an appointment or a hearing, and check what your employer or school must provide under the reasonable accommodation rules.
Caption glasses are a promising option, not a proven one. For the wider set of options, including built-in captions on phones and computers, see the guide to speech to text for deaf and hard of hearing people. If you need captions on calls rather than in person, read what a captioned telephone is. The full set of accessibility guides is in the accessibility and health overview.
Evidence
Every number on this page, with where it comes from. Open a fact to see the exact words in its source.
- 1.
- 2.
- 3.
- 4.
- 5.
- 6.
- 7.
- 8.
- 9.
Go deeper
Related questions
- Is automatic captioning accurate enough for deaf people?
- What is a captioned telephone?
- Can a pinched nerve in the neck feel like carpal tunnel?
- Can Access to Work pay for dictation software?
- Can ADHD affect a child's speech?
- Can carpal tunnel come back after surgery?
- Can carpal tunnel go away on its own?
- Can carpal tunnel syndrome be cured without surgery?