About the project:
- The app allows users to record their speech and will compute a score for the speech and provide feedback based on various metrics such as volume, speed, tone, and clarity
- By using machine learning models of different speech categories, the app is able to classify the quality of speech in real-time
The process:
- Developed the app frontend with
SwiftUI - Created machine learning models via
CreateML(CoreML) with over 90% validation accuracy for various speech metrics - Set up the backend to manage authentication and save user info with
Firebase
