Our SuperUROP project will attempt to address the problem of analyzing speech of speakers with speaking disorders by using learning techniques to characterize acoustic cues produced by these speakers. The acoustic cues are found in speech spectra and are reflective of the articulation of sound production in the vocal tract. Since these cues can be interpreted into evidence for distinctive features, phonemes, and words, we will develop software for finding the intended word sequence in the presence of speaking disorders.

“I am participating in SuperUROP in order to broaden my software experience as well as to learn more about linguistics as learning a foreign language has made it interesting for me. I hope to be able to learn new skills and provide what I already have to my team’s efforts.”