Voice recognition system with direction, translation and display

Project number: 
22025
Sponsor: 
McDonald/Watt Projects
Academic year: 
2021-2022
Many hard of hearing or deaf individuals depend on a translator or lip reading in group settings.

The Smart Voice Recognition System (SVRS) design helps users identify what a speaker is saying via voice-to-text, who the speaker is, and where they are located in a room – all using an iPad graphical user interface (GUI). Wireless transfer of direction information and audio files ensure ease of use and portability.

SVRS incorporates a neural network learning model to accomplish the voice recognition component. A built-in voice-to-text framework provides real-time mapping from speech to text. Filtered audio detected by the microphone array and defined by a GUI arrow identifies directionality. A Raspberry Pi connected to the microphone array houses all of the directionality software. The application that houses the voice recognition and voice-to-text software is loaded onto an iPad. And, the iPad and Raspberry Pi are connected via wireless ad-hoc network using MQTT.

Get started and sponsor a project now!

UA engineering students are ready to take your project from concept to reality.