Say, I pronounce a word and record it using a certain microphone. This record is treated as a master word. Then I record the same word again using the same microphone. I need to somehow compare the second sample to the master word and tell if it's the same word or not. I was thinking about a simple comparison with a threshold in percent. Say, > 80 % match means that it's most likely the same word, < 80 % match suggests that it's highly likely a different word.

It's important that this in no way should be any kind of speech recognition system, that would definitely be too complicated and an overkill in my situation. I'd rather call it a sound matching system

How should I go about this and what programming language/framework/library can work best for my goal?

  • 1
    $\begingroup$ This is in every way a speech recognition system, if perhaps the most simple form. So you can't make it "in no way a speech recognition system". It's like asking me to mix white, red, and a bit of blue paint for you, but by no means hand you a bucket of pink paint. $\endgroup$
    – TimWescott
    Feb 4 '21 at 20:50
  • $\begingroup$ Reading your comment the first time, I tended to completely agree. But then I realized that actually this actually is not supposed to be a "speech recognition system", but rather a sound comparison system. Thanks for pointing this out, I'm going to fix this misunderstanding in my text. $\endgroup$
    – Max Smith
    Feb 4 '21 at 21:17

Your Answer

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

Browse other questions tagged or ask your own question.