Public Safety and Crime Surveillance
What it does
Assists the Police Officers in capturing Real-Time data based on Public Crime Demographics and Computer Vision - keeping in mind Public Safety and police protection. The camera triggers as soon as it senses a danger based on sound, vision and local criminal history. Additionally, it keeps an eye on the surroundings to give an edge to Police Officers in crime scenes and collecting evidence.
How we built it
As a Team, we undertook the responsibility of individual tasks - connecting the pieces in the end together. We trained separate Deep Learning Models for Real-Time Disturbance Audio Detection and Computer Vision to detect Danger Index. We used Public Data of Crime Demographics to provide additional confidence to the decision of both Models. We integrated the application on Swift and traversed data through Firebase between different platforms.
Challenges we ran into
 Unavailability of Preprocessed Labelled Dataset or model.  Integrating Python scripts with swift to traverse data in Real-Time between IOS and computer. Coming up with the coolest name of our hack and domain.
Accomplishments that we're proud of
 Building a community focused hack to solve a significant problem towards public safety and crime surveillance. Being able to execute and deploy a full-fledged System of programs on an app in a short period of time.
What we learned
We learned to integrate Deep Learning models with real-life applications. We learned IOS development and the use of Firebase Platform. We learned to perform Data Augmentation on public datasets and derive mathematical formulae for calculations that analyzed predictions. We learned to have FUN!!!
What's next for Copture
We plan to proceed with the intended application of the project and improve the existing algorithm to cover a wide variety of use-cases.