Efficient Acoustic Front-End Processing for Tamil Speech Recognition using Modified GFCC Features
Variations between words will increase when the number of speakers try to utter the same word in different ways. Moreover, the speaker characteristics may vary due to the difference in the glottal waveforms, vocal tract lengths and speaking rates. All these variations can be eliminated by using the frequency warping technique which can assist to normalize the Vocal Tract Length (VTL).

