US grant creates AI expression recognition program for speech-challenged children

A bid to help children with communication disorders has netted some United States researchers a five-year, $20 million grant. The money will create speech recognition algorithms and fund the new AI Institute for Exceptional Education.

The competitive grant from the National Science Foundation went to the University of Buffalo and will focus on children with speech- or language-processing challenges, according to the school, which is part of the State Universities of New York system.

An AI Screener developed by the institute can collect pieces of a child’s speech, facial expressions and gestures for weekly summaries, which staff can use to monitor their abilities. They can also prompt a case to be passed on to a pathologist.

The AI Orchestrator is designed to act as an AI teaching assistant, offering interventions to children.

The school says there is a nationwide shortage of speech pathologists, resulting in insufficient services for affected children aged three to 10 years old.

“Children are not receiving life-changing interventions soon enough,” according to Venu Govindaraju, professor in UB’s computer science and engineering department.

Article Topics

AI  |  biometrics  |  biometrics research  |  expression recognition  |  gesture recognition  |  speech recognition  |  University at Buffalo