Project using ML and DS models for audio classification
Some of the results from the voice classification study were reported in this poster.
The subjects read a list of relatively similar sentences. Interestingly, some sentences demonstrated higher accuracy in speaker identification. Some speakers also had higher accuracy in the classification among different sentences.
It will be very interesting to dive deeper into the elements of the sentences and pronunciations to further analyze what could be the factors that led to higher accuracy in speaker identification.
