Inspiration
Many people get into trouble when drunk. And smartphone is an active part of that trouble. People can call, text, etc, and regret about that.
What it does
Automatically detects intoxicated behaviour and blocks undesired user activities.
How I built it
Mobile dev with Qt, audio classification with TensorFlow, accelerometer analysis with Fourier transformation.
Challenges I ran into
No free public datasets - need improvization
Accomplishments that I'm proud of
App
What I learned
Training neural nets for audio classification
What's next for Boohoscope
Publish in PlayStore, gather datasets, improve precision and recall, use smart bands
Log in or sign up for Devpost to join the conversation.