Term 252

Smart Selective-Hear Headset

Project Type: Self-Initiated

Project Description

This project presents the design and implementation of a Smart Selective-Hear Headset that enables users to prioritize desired voices and critical environmental sounds while suppressing unwanted background noise. The system captures ambient audio using multiple digital MEMS microphones and performs real-time processing on an embedded microcontroller with DSP capabilities, applying adaptive filtering, blind source separation, and active noise cancellation techniques. A software-based sound categorization module performs source identification and voice prioritization based on user input through a Bluetooth Low Energy mobile application, targeting high identification reliability in typical indoor environments and consistent voice isolation performance under multi speaker conditions. The headset is powered by a rechargeable lithium-ion battery designed to support several hours of continuous operation per charge, while thermo-adaptive control monitors internal temperature and humidity to regulate power consumption and ensure safe wearable operation. The integrated prototype demonstrates real-time selective hearing, reliable source separation, stable wireless control, and sustained battery-powered performance suitable for extended daily use