SpeechSense
The Project
The SpeechSense mobile application consists of three core components: a cross-platform Flutter app available on both major app stores, a Flask back-end server hosted on Render.com, and a C-Support Vector Classification model. The SVC model is trained on documentary audio for "safe" classifications and inebriated speech from Drunk History for "inebriated" classifications. Results are saved to Firebase, where connected hardware reads the status and prevents vehicle use if the user is deemed unsafe to drive.
