HackaWu

Inspiration

Our inspiration for developing SentryWu stemmed from a teamate perosnal experiance where he found himself in the middle of a robbery. The situation was overwhelming, and the response time felt agonizingly slow. This experience motivated us to create a tool that could detect potential threats in real-time, prevnet similiar situraion and have better response times.

What it does

The application, SentryWu, provides real-time detection of potential threats, such as weapons or suspicious actions, using advanced machine learning and computer visual algorithms. It continuoimmediately logs these events and provides alerts, helping to enhance response times in critical situationsnce of weapons. The system immediately logs these events and provides alerts, helping to enhance response times in critical situations.

How we built it

We trained models that can detect when people are violent and/or fighting among themselves, also we have models that detect a variety of weapons and a hand-gesture/body model that detects when people are raising their hands because they are feeling threatened.

Challenges we ran into

One of the major challenges we faced was incorporating all three models all together, we managed to merge some of them together but we did not manage to have them all three merged into one due to some dependencies problems, besides that we might've been able to make them all work together into one.

Accomplishments that we're proud of

They where several key accomplishments during the development of SentryWu. First, we successfully implemented a real-time detection system that uses machine learning to recognize weapons and suspicious behaviors. This was a challenging technical feat, requiring precise algorithm training and optimization. Another highlight was creating a user-friendly interface that provides clear insights while maintaining backend complexity. Lastly, our team collaboration, using GitHub to manage version control, ensuring efficient teamwork and code management.

What we learned

We learned that training models is not as easy as people make them look, they can sometimes be more challenging than one can imagine. Besides that we can eagerly say that we managed to accomplish everything we planned.

What's next for HackaWu

We would like to further tune HackaWu to be able to reduce the amount of false positives and false negatives so we can have a market ready product and implementation of Predictive Analytics and Threat Forecasting by working on incorporating predictive analytics that leverages historical data to forecast potential security breaches or threats, allowing users to take preventative measures before incidents occur.