-
-
Launch Screen
-
detect/report suspected DUI driver
-
share with law enforcement
-
confirm app works
Inspiration
After accessing the validated data, I realized I didn't have a continuous live stream but a snapshot of a moment in time. Therefore, I focused on video/image analysis.
What it does
The iOS app permits the user to select a specific time frame for analysis and determine if digital content capture by a node can determine a possible DUI situation. The meta data can be forwarded as a witness/victim statement to law enforcement for additional review and disposition. The choices are possible DUI driver, pedestrians only.
How I built it
Integrated IBM Bluemix to allow for machine learning. After obtaining the meta, created algorithm to distinguish be possible DUI drivers or pedestrians only.
Challenges I ran into
I would have preferred direct access to a continuous, live-stream, to create a more robust solution.
Accomplishments that I'm proud of
I got the app to work with UX/UI
What I learned
The data from platform is astounding and quite useful.
What's next for See Things.
If I get access to the live platform, create the app for real.
Log in or sign up for Devpost to join the conversation.