MIT EECS | Analog Devices Undergraduate Research and Innovation Scholar
Evaluating Acoustic Cue Labeling
- Natural Language and Speech Processing
Acoustic cues are the lynchpin between the physical and abstract domain of speech, an intermediate step between physical sound waves and full words. Everyone’ s use of acoustic cues are slightly different, giving them unique accents and inflections, but we still understand each other. My lab is working to quantitatively measure these differences to develop better accent detection, early speech disabilities detection, and speech synthesis. My project is to create an automated evaluation and feedback tool to train new lab members to understand and recognize acoustic cues from raw data. This will speed up the training process the lab currently has, and the code base will also provide the foundations for future work such as comparing accents.
I am participating in SuperUROP because I want to conduct research in a topic I find interesting. In the past, I’ve worked in research projects as a supporting developer, but through SuperUROP, I hope to learn more about writing papers and presenting research. I want to learn more about the subject matter of my project and about the research process. I’m most excited about analyzing the results of my research.