听障口语者语料库、语音信息库建设初探OA

Speech data from hearing-impaired speakers constitute a fundamental resource for research on hearing-impaired speech feature analysis and AI-driven voice restoration technologies.China has a large population of hearing-impaired individuals,among whom many use spoken language as their primary means of communication.However,dedicated speech data for Mandarin-speaking hearing-impaired individuals have long been lacking in quantity,standardized norms,and sharing mechanisms,which has become a critical bottleneck hindering the deployment of AI-based rehabilitation technologies.In response to the current lack of systematic standards for hearing-impaired speech data construction,this paper reviews relevant theories and practices from both domestic and international sources,and analyzes the localization challenges faced in China.On this basis,this paper proposes leveraging artificial intelligence technologies to promote the systematic,standardized,and intelligent construction of hearing-impaired speech databases through the establishment of industry standards,the development of human-AI collaborative workflows,the creation of tiered open data sharing mechanisms,and the acceleration of interdisciplinary talent cultivation.This study aims to provide methodological references for the systematic construction of speech data resources for hearing-impaired individuals in China.