Inspiration
Defunctory Emergency sytems in India
What it does
Filters the calls based on graveness for the memergency services operators because sometimes every second counts. The same model can also be used in other applications like SOS applications on phones that will trigger SOS based on audio detected and not wait for us to press 20 buttons to send them which you would have the time to do in a dangerous situation anyway.
How we built it
Usind speech emotion dtection based on liberosa and python and other supposrting libraries
Challenges we ran into
The dataset smapling rate was too high and took forever to down sample it.
Accomplishments that we're proud of
The accuracy is pretty good for initial dev.
What we learned
Better dataset and model is required
What's next for Emotion detection system for SOS message filtering
UI, end to end service provide and integration with emergency contact frameworks as part of smart city initiatives.
Log in or sign up for Devpost to join the conversation.