Speech recognition using MFCC and DTW
Citations
64 citations
51 citations
Cites methods from "Speech recognition using MFCC and D..."
...Lastly, we use the log-transformation of amplitude values to normalize (addressed as log normalization) and emphasize on low intensity components (inspired by speech recognition literature [52, 57, 59])....
[...]
47 citations
45 citations
Additional excerpts
...Words recognition to compare the similarity of letters in words “ forward”, “Left”, “Right”, “Reverse”, and “Control” [4]....
[...]
33 citations
Cites background or methods from "Speech recognition using MFCC and D..."
...Some studies concerning speech recognition system using MFCC as the feature extraction methods have been conducted [3]–[6]....
[...]
...with N is the number of samples per frame, Y[n] is the output signal, X[n] is the input signal, and W[n] is the nth coefficient of the Hamming window [3]....
[...]
References
21,819 citations
846 citations
"Speech recognition using MFCC and D..." refers background or methods in this paper
...DCT is calculated using equation shown in (6) [6]....
[...]
...Features obtained by MFCC algorithm are similar to known variation of the human cochlea’s critical bandwidth with frequency [6]....
[...]
...On the other hand it should not be to too long such that under a particular frame voice sample is time invariant [1,6]....
[...]
...For simple isolated word detection MFCC and DTW approach is enough and efficient [6]....
[...]
...DTW finds the optimal alignment between two times series if one time series may be “warped” non-linearly by stretching or shrinking it along its time axis [6]....
[...]
281 citations
"Speech recognition using MFCC and D..." refers background in this paper
...The feature matching algorithm cannot discern the difference between two closely spaced frequencies [9]....
[...]
63 citations
"Speech recognition using MFCC and D..." refers background in this paper
...The equation for calculating MEL for a given frequency is shown in (5) [8]....
[...]
...Pre-emphasis stage increases the magnitude of higher frequency with respect to lower frequencies [8]....
[...]
...They avoid interaction of noise with significant features [8]....
[...]
42 citations
"Speech recognition using MFCC and D..." refers background in this paper
...Combination of various features is to be adapted in this case for high reliability [4]....
[...]
...Various methodologies have been proposed for isolated word detection and continuous speech recognition over the years [4]....
[...]