This study presents the development of a voice-enabled smart personal assistant to support instructional activities within smart classroom environments. Distinct from prior research that has primarily focused on desktop-based or mobile implementations, this work integrates the Whisper speech-to-text model into a client-server architecture, utilizing the Jetson Nano as a low-power computing node. The system is designed to facilitate hands-free interaction and adaptive control of classroom operations. Functional evaluations demonstrate a command recognition accuracy of 95%, an average inference time of 3.5 seconds, and an effective operational range of up to 3 meters. These results highlight the feasibility of deploying lightweight, voice-driven AI solutions that uniquely address limitations in low-resource educational environments..
Copyrights © 2025