Speech recognition

Siri, Alexa, “Hey Google” — computers are getting better and better at understanding our voice. Hypersay is bringing the power of speech recognition for the benefit of your participants.

We are using Google’s services to automatically transcribe as captions (subtitles) the spoken words in the language of your choice. Languages recognised: 64 (120 if you include accent variations).

“We can’t hear you!”, “What did she say?”, “Oh, I missed that” — these will no longer happen. Now your participants have the transcription both on the large screen and in their connected devices.

Live translations

Let’s push things a bit further. Now that we have speech recognition working for your presentation, the participants can opt to change the destination language to something they are more familiar with.

We use the same trusted Google services to automatically transcribe into 105 languages. We are confident your preferred language is among those.

Accuracy

“Lost in translation” is not just the name of a great movie, it is also a reality of a globalised world. Computers do their best to use Ai, rich context and complex algorithms to automatically recognise speech and perform real-time translations. But they will fail, from time to time, so accuracy becomes a problem.

Factors such as background noise, quality of microphone, local accent, internet speed and pace of speech will highly influence the number of mistakes that occur.

However, if the audience does not speak your native language, they will benefit from understanding 80% of what you say. Speak slowly on important points, repeat them if necessary and your message will be understood.

Important tip: use your usual wireless Apple AirPods, Samsung Galaxy Buds, Echo Buds or any other preferred wireless earbuds. They will greatly improve the quality of speech recognition.