Voice-Activated Assistance for the Elderly: Integrating Speech Recognition and IoT
Narumol Chumuang, Thittaporn Ganokratanaa, Patiyuth Pramkeaw, Mahasak Ketcham, Suvil Chomchaiya, Worawut Yimyam
Abstract
This work aims to develop an elderly assistance system using voice commands by applying speech recognition technology. The system is designed to enhance convenience for elderly individuals and also reduce labor costs associated with hiring caregivers. It leverages the principles of the Internet of Things (IoT) to control various household electrical devices through voice commands. Through Natural Language Processing (NLP), the system processes natural language to interpret voice commands, allowing users to control lighting, open applications, make phone calls, and more. The researchers utilized a Confusion Matrix model to measure the accuracy of the voice command recognition system. It assessed whether the system correctly interpreted the voice commands issued by users. The testing, conducted with a group of individuals aged 55-60, revealed an accuracy rate of 85%. This demonstrates the system’s effectiveness and potential to augment the independence of elderly individuals. In essence, this work presents a voice-command assistance tool that combines speech recognition and IoT, promising enhanced convenience for the elderly and potential savings in caregiving expenses.