We want to address the "elephant" in the room, pertaining to situations of verbal abuse and discrimination.
What it does
Provides speech sentiment analysis for potentially charged, sexist, discriminatory speech. Anything that falls into the parameters of "isms"
How we built it
We deployed an iOS and web app built on X-code sdk and node.js respectively. Web app uses web audio api plug in for speech recording data, node.js backend with express, and api integrations.
Both apps post to back end running a Flask API that passes sound data through a Tensorflow speech-to-text algorithm, runs the parsed text through sentiment and text analysis, and then returns results to the apps.
Challenges we ran into
Posting requests with audio content in blob format and getting accurate sentiment detection with the NLP library.
Accomplishments that we're proud of
Both apps are working! And added additional features. Worked with an excellent team :)