0
$\begingroup$

I want to build a Automatic Speech Recognition (ASR) engine for myself, but I've no idea from where to start.

I've read that most ASR's are build upon Hidden Markov Models, but also I've read that HMM is limited somehow and a better approach is to build a ASR upon Machine Learning features.

Overall I am confused.

Can someone recommend me books on topics from which I can start?

Thanks (:

$\endgroup$

closed as too broad by D.W., David Richerby, Luke Mathieson, Juho, jmite Feb 2 '15 at 9:10

Please edit the question to limit it to a specific problem with enough detail to identify an adequate answer. Avoid asking multiple distinct questions at once. See the How to Ask page for help clarifying this question. If this question can be reworded to fit the rules in the help center, please edit the question.

  • 3
    $\begingroup$ 1. This question is too broad and you're probably better off spending more time self-studying standard resources on speech recognition. There's lots written about it out there, so if you've been having trouble finding resources, you haven't done enough searching. 2. Book recommendations are generally off-topic for this site. See meta.cs.stackexchange.com/q/20/755 and meta.cs.stackexchange.com/q/874/755. $\endgroup$ – D.W. Feb 1 '15 at 19:33
  • $\begingroup$ What is an ASR for you? HMM or equvalent techniques are only the very beginning of a long processing chain. How far do you want it to go? $\endgroup$ – babou Feb 1 '15 at 20:16
  • $\begingroup$ I want to build a simple ASR which recognizes only few words ("yes" and "no" for example) $\endgroup$ – 0xdeadcode Feb 2 '15 at 0:05
3
$\begingroup$

I want to build a Automatic Speech Recognition (ASR) engine for myself, but I've no idea from where to start.

Start with trying existing open source speech recognition system, learn how they work, play with them. Check http://cmusphinx.sourceforge.net.

I've read that most ASR's are build upon Hidden Markov Models, but also I've read that HMM is limited somehow and a better approach is to build a ASR upon Machine Learning features.

HMM which provide time variance is still a reasonable model for speech, most of the system use it in one form or another. Limitations of HMM exist, but you do not need to care about them on that level.

You can implement a basic speech recognition system without HMM. For example check

http://de.mathworks.com/company/newsletters/articles/developing-an-isolated-word-recognition-system-in-matlab.html

That would be enough for "yes" and "no" discrimination. For any more serious use you will need HMM.

Can someone recommend me books on topics from which I can start?

There are two good books on topic:

A basic guide to start:

http://www.amazon.com/Spoken-Language-Processing-Algorithm-Development/dp/0130226165

A modern state of the art

http://www.amazon.com/Automatic-Speech-Recognition-Communication-Technology/dp/1447157788/

$\endgroup$

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