The goal of my project is to help complete the development of an automatic speech analysis system, making use of what is known about how humans process and understand speech. Potential applications include automatic speech recognition, diagnosis of clinical speech disorders, and a better understanding of human speech production and perception in general. The words that peo- ple speak are formed from sequences of phonemes, each of which can be specified by several discrete features (e.g., “nasal”) that are in turn are characterized by certain acoustic cues in the speech signal. My part is to consolidate existing modules for detecting certain acoustic cues from a supplied speech signal. From there, phonemes and then words can be extracted.

I hope to learn more about speech acoustics, speech recognition, and the research enterprise.