Maria Ximena Rueda-Guerrero

Maria Ximena Rueda-Guerrero

Scholar Title

MIT EECS Undergraduate Research and Innovation Scholar

Research Title

Specific-Language Impairment Diagnosis Based on Landmark Modifications in Non-Word Repetition Tasks




Electrical Engineering and Computer Science

Research Areas
  • Machine Learning

Stefanie Shattuck-Hufnagel


Automatic analysis of speech from typically and atypically developing children
In this project we will analyze typically and atypically developing children’s speech using the landmark/acoustic cue system developed by the Speech Communication Group. Landmarks are robustly-detectable locations in the signal that indicate classes of speech sounds. Additional acoustic cues near landmarks specify other distinctive features related to place of articulation and voicing that define each sound. The realizations of landmarks and other acoustic cues are strongly impacted by prosodic events such as phrase boundaries and phrase-level accents. We have applied this feature-cue-based approach to child speech elicited by non-word repetition tasks with hand labeling of the landmarks and the next steps will be to develop a software for automatic detection of landmarks.


I decided to participate in SuperUROP to engage in a long research project while also receiving guidance in performing research. I had participated in this same project as a UROP student and I wanted to continue it through the SuperUROP program too. I hope to learn more about linguistics and applications of machine learning for speech recognition algorithms.

Back to Scholars