Hello!

This app lets users record audio of themselves speaking various sentences that they can choose or are predefined in various languages. Then utilizing Google's Speech-to-Text API and our own algorithms which analyze and recognize mistakes in the user's pronounciation and enunciation in a language generic fashion and returns to the user sample audio to correct their mistakes. This app is meant to help people improve their communication skills and confidence while speaking and can prove useful for a variety of groups including immigrants, second language learners and children.

Currently the speech recognition supports over 35 different languages, and supports 10 of those languages for sample audio given back to the user.

Technologies Used: Google Cloud Speech-to-Text API Web text-to-speech API Node JS Express JS EJS Mongo DB RecroderJS

NOTE: to get this app working properly, you need a valid GCP acocunt key

References: Text to speech help: https://github.com/Nitij/JavaScriptSpeechSynthesis/blob/master/index.html

Pulse animation: https://www.florin-pop.com/blog/2019/03/css-pulse-effect/

Recording user audio from borwser: https://github.com/mattdiamond/Recorderjs https://github.com/addpipe/simple-recorderjs-demo

Contributers: Husni Fareed, Muhammad Saad Raza, Mustafa Hafeez