AICanHear: Assistive Alerting Device for Deaf and Hard-of-Hearing People (DHH)

Hearing loss is a significant global health issue, affecting an estimated 1 billion young people at risk of irreversible damage. To address this, the researchers developed "AICanHear," a wearable Sound Classifier and Locator device for the Deaf and Hard-of-Hearing (DHH) community. Two prototypes, a wristwatch, and headband, were developed, each utilizing specific hardware and software components. To achieve sound classification, the study used YAMNet, a pre-trained sound classification model accessible via the web. The experiments revealed AICanHear device’s accuracy of 70.48% indoors and 63.33% outdoors in classifying sound events. Sound localization accuracy varied across directions, ranging from 36.67% to 56.67%. User feedback indicated positive acceptance and satisfaction, reflecting the device’s effectiveness in improving communication and safety.