Many times, students are told to practice their speeches in front of a mirror. We want to make a better mirror.
What it does
Listens to your speech and provides you with invaluable feedback on your speech. This includes how fast you speak over time, stutters and duplicated words, and an emotional measure of your project
How we built it
Our stack is: React JS provides our listening service and graphing tools The Flask Python framework performs speech to text services, calculates various metrics, and also communicates with the IBM Bluemix Natural Language Processing API
What we learned
Audio encoding isn't that fun.
What's next for Speak
More data, more analytics. We want to run our tool on major audio data sets and fine tune our own natural language processing model! We can use our data to improve our suggestions.