Secure Sight

Inspiration

In times of emergency, every second counts, and unfortunately, not all victims can reach out safely or quickly enough to call for help. This inspired us to create Secure Sight, a proactive and accessible security solution that leverages machine learning to detect potential threats in real-time and alert security personnel. We wanted to build a tool that empowers individuals to anonymously report incidents with supporting evidence, bridging the gap between threat detection and law enforcement.

What it does

Secure Sight uses a machine learning model to analyze live CCTV camera feeds, detecting the presence of weapons in real-time. If a weapon is detected:

The system uploads the evidence to an Amazon S3 bucket to maintain data integrity.
Automatically sends an alert to security admins via WhatsApp using the Twilio API. The alert message includes detection details, location, and an image for validation.
Features a web interface for anonymous crime reporting, allowing users to submit incident descriptions and evidence without personal identification, encouraging community involvement without fear of scrutiny.

How we built it

We built Secure Sight using a blend of machine learning and web technologies:

Machine Learning Model: Processes CCTV feeds for real-time weapon detection. Fine-tuned for accuracy in diverse scenarios.
AWS S3 for Storage: Ensures secure and accessible storage of detected evidence.
Twilio API for Notifications: Sends alerts via WhatsApp to admins without needing additional apps.
Web Interface with Vanilla JavaScript: Provides a lightweight and user-friendly interface, including an anonymous reporting system.

Challenges we ran into

Building Secure Sight presented several challenges:

Model Accuracy: Initially struggled to distinguish weapons from other objects. Fine-tuning and testing improved reliability.
API Integration with Vanilla JavaScript: Uploading images to S3 and implementing Twilio API with JavaScript was complex and required multiple refinements.
Anonymous Reporting Feature: Ensuring user privacy while maintaining data reliability demanded meticulous design.

Accomplishments that we're proud of

Built a comprehensive, end-to-end solution that takes immediate action by alerting admins via WhatsApp.
Developed an anonymous reporting feature to encourage discreet reporting of incidents.
Created a user-friendly web interface and integrated secure, scalable cloud storage.

What we learned

AWS S3: Efficient data storage and retrieval techniques.
Twilio API: Building WhatsApp notifications and managing API requests effectively.
Machine Learning for Real-World Applications: Optimizing models for real-time analysis in diverse conditions.

What's next for Secure Sight

Further improve detection accuracy.
Optimize the solution for edge devices (low-power systems/IoT).
Expand deployment to schools, businesses, and public spaces to make Secure Sight an affordable, proactive security tool for crime prevention.