Digital assistants like Siri and Cortana have trouble recognizing proper nouns and programmers have to spend hours hard coding in the phonetics for names that cannot be predicted algorithmically.
What it does
Our app is a social game in which one person tries to construct a sentence to make the other user laugh. In addition to being an entertaining and addicting game, it contributes to the advancement of voice recognition technology by collecting and associating text and speech clips of hard-to-predict proper nouns.
How we built it
We used PostgreSQL for the database, Python, Flask-ReSTful, ORM, and Microsoft Azure for the backend and Android and iOS Swift apps for the frontend.
Challenges we ran into
The iOS app had difficulty connecting to the API we made, especially since the Android and iOS audio formats were incompatible. We started later than we had anticipated because we struggled to come up with a game that would make people want to speak proper nouns into their phones.
Accomplishments that we're proud of
Our backend functions and database calls are all usable! We have a fully functional Android app and mostly functioning iOS app despite early struggles.
What we learned
We learned about 5 different ways to play an audio file in an iOS app and how to divide and conquer effectively.
What's next for LaughOutLoud
We scale up our apps and sell the data to bigger companies.