AI-powered Speech Therapy System Prototype

This AI-powered speech therapy system enables speech therapists to create personalized training content and generate realistic model speech for each practice session. During training, a virtual therapist presents the target phrase and plays the model speech for the patient to hear before attempting the phrase.

The system records the patient’s response and uses automatic speech recognition to generate a transcription and performance score. These results allow patients to practise repeatedly while helping therapists review and monitor their progress.

01
Personalized Content and Model Speech
02
Virtual Therapist-guided Practice
03
ASR Transcription and Scoring

Demonstration