I would like to know the steps in making a java speech recognizer/API WITHOUT using sphinx or anything else. I want to build from the ground up!!! Also a Text to speech from the ground up. 100% my own. I know it's going to be a lot of work. I just want to know.
Asked
Active
Viewed 1,465 times
0
-
3And what have you found so far from your own research? – Thorbjørn Ravn Andersen Sep 01 '14 at 07:53
-
Everything says use sphinx but I want my own thing I want to know how – Kenny Hawkins Sep 01 '14 at 07:55
-
3This require advanced math and programming knowledge. And even with those, this will not be an easy task. Anyway, I suggest that you go through the research section of the Sphinx project: http://cmusphinx.sourceforge.net/wiki/research/ – ortis Sep 01 '14 at 07:58
-
"I want to know how". That is what your own research should help you find out. – Thorbjørn Ravn Andersen Sep 01 '14 at 09:12
1 Answers
4
First you want your sound-input as bytestream, how to do so is explained by oracle:
http://docs.oracle.com/javase/tutorial/sound/capturing.html
Next, you write the logic and algorithms to analyze and process the byte stream.
http://en.wikipedia.org/wiki/Speech_recognition#Algorithms
Good luck.
ASA
- 1,911
- 3
- 20
- 37