Skip to content

ishaanjav/TrafficAI

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

18 Commits
 
 
 
 
 
 
 
 
 
 

Repository files navigation

TrafficAI

Inspiration

According to the Federal Highway Administration, there is an average of 6 million car accidents in the United States alone. Approximately 6% or 36,000 of those accidents will result in at least 1 death. 25% of these deaths could have been prevented had the police and ambulance been called earlier by anyone. At the moment, the average time to call emergency services is between three to five minutes. These accidents affect not only the directly injured and their close ones but also Americans as a whole. More accident-related injuries result in an overall increase in healthcare costs, which the average American has to bear due to insurance. So, a reduction in the severity of injuries would be of benefit to everyone.

What it does

TrafficAI is an instant car crash detection program. We used machine learning to train a model to detect normal traffic and car crash videos and deployed it onto a website. From CCTV footage, the model predicts if there is a car crash in each frame, and when there is, the website audibly alerts a security guard and provides them with real-time footage of the crash from traffic cameras as well as the location. The guard also has a button they can easily click to alert authorities once the crash has been confirmed.

About

TrafficAI: AI-based software detection of car crashes through the analysis of spatiotemporal features of a sequence of image frames over time.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors