how to separate words from a sentence

Question

4.67/5 (3 votes)

See more:

suppose I have a speech signal (lets say, of 5 seconds duration). The words are isolated i.e. words are separated by brief pauses, but total number of words are unknown in the speech signal. How can I separate these words? For the argument sake, lets say there are 5 words in the signal, how can I separate these words as well as store them in 5 different variables?
Thanks in advance.

Posted 25-Feb-11 9:34am

Member 7702812

Add a Solution

2 solutions

Add a Solution

Add your solution here

Treat my content as plain text, not as HTML

Preview 0

…

Existing Members

Sign in to your account

...or Join us

Download, Vote, Comment, Publish.

Your Email
Password
Forgot your password?

Your Email
This email is in use. Do you need your password?
Optional Password

I have read and agree to the Terms of Service and Privacy Policy
Please subscribe me to the CodeProject newsletters

When answering a question please:

Read the question carefully.
Understand that English isn't everyone's first language so be lenient of bad spelling and grammar.
If a question is poorly phrased then either ask for clarification, ignore it, or edit the question and fix the problem. Insults are not welcome.
Don't tell someone to read the manual. Chances are they have and don't get it. Provide an answer or move on to the next question.

Let's work to help developers, not make them feel stupid.

This content, along with any associated source code and files, is licensed under The Code Project Open License (CPOL)

Viktor Signaievskyi · Answer 1 · 2011-11-20T08:44:00

Hi Member 7702812,

I think I do understand your question:)

Funny topic and so much solutions...:(

The answer is:
"Set imaginary pointer to the beginning of the initial samples. Take n samples from initial set of N (where n much less than N) samples and find the power for it (f.e. summarize abs values). Then shift your imaginary pointer to some shift and also take n samples and do the same... and so on...
The end of the word will be in the place where previous power will be greater by some threshold relatively to the current power of the signal."

I think the idea is clear to you. If no - I can provide you by code snippet...

God luck!

Sandeep Mewara · Answer 2 · 2011-02-25T19:32:00

Solution 1

I went into Google search and did this[^] and got the first link as:
Similar discussion[^]

Second link as: MATLAB Split[^]
I guess, you would be able to move ahead now.

Posted 25-Feb-11 19:32pm

Sandeep Mewara

Comments

Member 7702812 26-Feb-11 2:49am

you didn't understand my question........I have a SPEECH SIGNAL, not a text sentence.........Matlab doesn't even know that the signal is a sentence

Sandeep Mewara 26-Feb-11 2:50am

Try Speech to Text and then move ahead. What say?

Sandeep Mewara 26-Feb-11 2:51am

So, you are saying you tried. Googled and all, and you did not find anything? What I said was known to you?

Member 7702812 26-Feb-11 2:53am

yes I have tried a lot, what you have given is definitely a solution, but not related to my problem, its totally different.
Anyway, thanks a lot for you time

Sandeep Mewara 26-Feb-11 2:56am

Ok. I take it as a true comment. Take a 5 for effort made in trying.

Member 7702812 26-Feb-11 3:03am

haha, thanks a lot man
I guess words isolation is still a major hurdle in automatic speech recognition
anyway, thankyou for your time