Hello everybody,
I have to reduce the noise of my speech signal. Are there any tested algorithms, written code or implemented programs that I can use within my application to speed up my coding duration?
I will use it within my thesis. My thesis is on speech recognition, but before the feature extraction process, I have to reduce the noise of the signal.
the main thing is bandpass filtering which removes noise outside the frequency range of interest.... next there is a need for echo cancellation which is due to reflection and other imperfections.... there are many algorithms for echo cancellation based MMSE, ML etc.... the other inherent noises are hard to reduce and are mostly due to channel....
To clarify ebook123456789's question, what kind of noise is it? What does it sound like?
If you can't identify the noise, maybe upload a small sample file that we can play.
If you attach the file to an EDAboard message, you'll need to ZIP it or RAR it first.