Fear of public speaking is one of the most prominent phobias in the world. Although so many people suffer on a regular basis because of this fear, little has been done to combat it.
What it does
Enter SpeakNLearn. We understand that most of the anxiety surrounding speeches comes from the fear that one might forget a point or two, and our app seeks to stop that. Our app helps people practice wherever and whenever they want to, and gives them pointers about accuracy and timing to make their speech truly remarkable.
How we built it
At the start of the app, we use the Android Speech-To-Text API for voice recognition to track the speaker's progress and accuracy as they move on. At the end of the speech, we use the Google Cloud Natural Language API, along with a homebrew cosine-similarity NLP analysis, to gauge the accuracy of the speech and generate a few pointers where the speaker can improve.
Challenges we ran into
We had trouble integrating the Google Cloud Natural Language API at first, which motivated us to improvise our own method. Once we got the API to work, we used the two measures to determine an even more accurate measure, as well as certain points which the speaker missed.
Accomplishments that we're proud of
We are proud of making an app from scratch in an overnight hackathon, as well as integrating two APIs into our app to create a truly immersive experience.
What we learned
We learned how to integrate APIs into our Android apps.
What's next for SpeakNLearn
We hope to expand the horizons of our app to include Google Play publication and Google Home integration.