Skip to content

THE INTEGRAT AI PROBLEM SET PHONEME RECOGNITION

Professor Nussbaum edited this page Mar 27, 2018 · 1 revision

THE INTEGRAT AI PROBLEM SET - PHONEME RECOGNITION

All spoken languages are made up of phonemes. Phonemes are the linguistic elements, that when changed, can change one spoken word into another. All speech recognition algorithms begin with phoneme recognition, and then determine the "most likely word" that was spoken, based on the sequence of phonemes detected.

All humans who can speak and understand a particular language fluently, are by definition, experts at phoneme recognition.

For this reason, INTEGRAT focuses on phoneme recognition. By choosing this AI problem set, it is nearly guaranteed that the AI developer is also an expert decision maker (when listening to speech in their native language, and deciding what was spoken).