1
$\begingroup$

How much training data is needed to build a speech-to-text engine based on machine learning? (To within an order of magnitude or so.)

Big companies like Google, Facebook have a massive amount of data. For usual people its not possible to acquire that amount of data and without it training a model for the purpose seems impossible right now. Is it possible for a few individual to build a speech to text recognition model comparable to Google?

$\endgroup$
  • 2
    $\begingroup$ "Can a few people build some piece of software?" doesn't seem to be a question about computer science. $\endgroup$ – David Richerby Jul 3 '17 at 7:10
  • $\begingroup$ Where should I post it @DavidRicherby. Thanks. $\endgroup$ – Tahlil Jul 3 '17 at 7:24
  • $\begingroup$ Modern machine learning methods rely on massive amounts of data and massive computation power. If you are not in possession of both, you will not do as good as Google or Facebook. $\endgroup$ – Yuval Filmus Jul 3 '17 at 7:42
  • $\begingroup$ @DavidRicherby Indeed, but there are two parts to this question: estimating the required volume of data, and how to obtain that much data. The first part is on-topic. $\endgroup$ – Gilles 'SO- stop being evil' Jul 3 '17 at 13:48
  • $\begingroup$ As much as Yahoo, Google, Apple, and other giant money-makers have. $\endgroup$ – fade2black Jul 3 '17 at 14:03

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.