Word comparison and Identification.

Good day all, I'm trying to compare words and find which word is most likely to have been said usaing the immse() (mean square error) function. I record 5 words which I need to compare to one word. (CWord). I've tried using psectrum and fft functions, obtaining the peak points with findpeaks() and multiple other ways of accomplishing this. Can anyone give me a clear path to take to achieve wwhat I'm looking for. Code would be nice but if you've done comparing audio files before and know what route to take that wwould be much appreciated. I've attached the words I want to compare to and the word I compare with (CWord). This is for an assignment and all my attempts have failed, so I'm quite despirite. Thanks in advance!

Answer 1

[Sahil Jain]

Hi. You can try computing the mean-squared error on audio features extracted using "audioFeatureExtractor" for better results. You can also refer to the answers on this question for more ideas.