AIs like Siri and Alexa respond to their names being called. How does the system recognize the name by ignoring all the other words that have been said before their name? For example, "Hey Siri" would trigger Siri to start listening for commands, but if a user said "hey how are you hey Siri" the system will ignore "hey how are you" but trigger the system to "hey Siri". Is it because their listening function reloads in milliseconds or even nanoseconds, or is there a different way it works?


Is it because their listening function reloads in milliseconds or even nanoseconds

Yes, it expects the keyword to start every moment of time and it ignores the rest.

Overall, the algorithm is described here, you can read for details:


| improve this answer | |
  • 1
    $\begingroup$ Fun fact: some cheaper Android phones don't have a dedicated audio processor to listen for the equivalent "Hey Google" or "Ok google", so it does cost significant power to enable always-on listening for that phrase. (Android makes it optional.) But yes, having a unique phrase of more than 2 syllables before firing up the full power of speech-to-text processing is important for saving power, as well as minimizing false positives so you can talk about siri with another human without your phone jumping in! $\endgroup$ – Peter Cordes Jul 26 at 21:31

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.