EL PASO, TX – The University of Texas at El Paso (UTEP) has received part of a grant from the National Science Foundation to conduct research in support of artificial intelligence, with the intention of aiding speech and language development in young children.
The five-year, $20 million grant will be used to create the Artificial Intelligence Institute for Exceptional Education, which will be run by the University at Buffalo and will address speech and language development problems faced by children between the ages of three and 10.
UTEP reported that the institute will consist of more than 30 researchers from nine universities, including UTEP; the University of Washington; Cornell University; the University of Nevada, Reno; Pennsylvania State University; the University of Oregon; and the University of Illinois Urbana-Champaign.
“Although speech processing technology works very well for adults today, there is still a long way to go. Most important for children is the ability to interact with peers and teachers. Effectively modeling that interactive behavior, and how it differs between typically developing children and those facing challenges, is the problem we will tackle,” said Nigel Ward, professor of computer science and head of the UTEP team.
The institute will work to develop two tools: the AI Screener and the AI Orchestrator. The former allows children to be listened to and observed in the classroom, creating weekly summaries of their speech, expressions and gestures. This data is used to suggest, if necessary, an evaluation with a speech therapist.
The AI Orchestrator, is an application to help speech-language pathologists with large caseloads – a challenge exacerbated by the Covid-19 pandemic – by providing customized content for each child they serve. All work must conform to ethical and privacy guidelines. The prototypes will first be used in about 80 classrooms, testing nearly 500 kindergarten students.
UTEP’s computer science department will receive $475,000 to contribute to the research. Ward and assistant professor Diego Aguirre’s research will focus on computational modeling, which can support automated detection systems such as AI Screener. Specifically, their work will model interaction patterns such as intonation, timing and rhythm.
Ward and Aguirre, along with Jonathan Avila, a student in the PhD program in computer science, will model these interaction patterns, with an emphasis on computational methods.