Risk Zone Image and Streaming Assessment

We developed an App which can be used from any platform through a webpage. It allows to upload an image or a video streaming to be analyzed for hazard zones and survivors.

Comment

Inspiration

Places of destruction are also dangerous for first responders which need a fast and reliable way to analyze the situation to safely perform their work.

What it does

The application accepts image and video streaming to and displays an analysis of the user's choice based on the what an AI observes in the image or video.

How we built it

The backend of the app is made in python and the frontend is built with HTML, Javascript and CSS. It utilizes advanced models to analyze the input.

Challenges we ran into

We had to developed an app around what a first responded would have at the time of the disaster. It had to be intuitive, minimal and be fast to make a response.

Accomplishments that we're proud of

At the beginning, the group thought that the used prompts for the app would need to be a lot for a first responder so we are proud of how fast an image can get analyzed which is perfect for when time makes a difference between life and death.

What we learned

This is the first time working with videos in an application project. We learned how much tokens the AI model would use for a streaming video and continuous feedback.

What's next for Risk Zone Image and Streaming Assessment

  • Since the app accepts video streaming. We could add options for continuous feedback from the AI to emphasize risks.
  • The photo side of the app. We could have the photo modified to better pinpoint areas of interest.
  • We could have a data bank with past rescues and places of destructions. Group the input with any past experiences so the AI would be better at suggesting certain actions.
Share this project:

Updates