As you may or may not know we use phonetics based speech analytics [as opposed to LVCSR ‘Large Vocabulary Continuous Speech Recognition’]. The reason we use phonemes is that words that are not in a predefined vocabulary can still be found, provided the phonemes are recognisable.
Phonemes are like ‘units’ of sound that create meaning! T OP ST OP DR OP RO OLL etc….
Recently Nat and Lo went into google to learn about speech synthesis and PHONEMES!
Check out this excellent behind the scenes on speech synthesis
Phonemes and Speech
To learn more about our managed speech analytics solution head over to http://www.knowledgespace.com.au