Hello,
Speech Recognition systems, rather than being a single
algorithm, actually involve a number of processing
stages and diverse algorithms. In particular, one can
describe the feature extraction stage, where features
useful for recognition are extracted from speech
signal, and the comparison/decision stages, where you
use the features extracted to compare with some
previously captured features and then decide on what
has been uttered. You need to decide on feature types
you which to extract such as MFCC LPCC etc, and the
decision as well as comparison algorithms (Vector
Quantisation, DTW, HMM, k-NN etc) to use.
Hope this helps
Olutope Foluso Omogbenigun
--- s...@yahoo.co.in wrote:
> Hi
> i am doing a project on speech recognition security
> system using DSP
> can any body help me to understand speech
> recognition algorithm
>
> i want to know how to do FFT with 1024 point and sum
> up first 128 coefficients
>
> please help
>