Open, Privacy Enhanced Activity Recognition
Let's create a safer world where public surveillance empowers communities without compromising individual privacy.
How It Works
Privacy First
The system activates only for likely suspicious activity, processing only relevant events to protect privacy and avoid unnecessary surveillance.
Real-time Detection
Achieves 93% accuracy on the UCF Crime Dataset, detecting events like Abuse, Arrest, Arson, Assault, Burglary, Explosion, Fighting, Normal Videos, Road Accidents, Robbery, Shooting, Shoplifting, Stealing and Vandalism with immediate local processing and alerts.
Open Source
Fully transparent hardware and software platform for community verification and enhancement.
Our Roadmap
Leading Crime Detection
- Trained and deployed a model achieving 93% accuracy on the UCF Crime Dataset
- Designed the system to ensure all data remains private within the platform
- Check out the model
Segmentation and Extraction
- Integrating a segmentation algorithm to identify and extract data on individuals related to the detected crime
- Ensuring anonymization for individuals not involved in the action
Intercommunication Protocol
- Developing a protocol for intercommunication between devices in proximity to the crime scene
- Testing and optimizing networked device functionality for real-time response
Integrated Hardware Platform
- Building a hardware platform compatible with existing CCTV infrastructure
- Ensuring seamless integration with public and private security systems
Government Collaboration
- Establishing partnerships with local governments to enable real-time crime reporting to law enforcement
- Adapting the platform to comply with local privacy and legal regulations
Transparency and Trust
- Open-sourcing hardware and software to foster community collaboration
- Publishing methodologies and ethical guidelines to build public trust
Our Vision
To create a safer world where public surveillance empowers communities without compromising individual privacy.
Our Mission
To develop cutting-edge, privacy-first technology that detects and prevents crime while protecting the anonymity of those unrelated to any actions. We aim to open-source our platform to foster collaboration and trust globally.