Oosto Video Analytics for the Real World

What is Oosto?

Oosto delivers advanced Vision AI technology designed for real-world security and safety applications. Its platform provides accurate, real-time insights by analyzing video feeds, enabling functions like person of interest alerting, smart behavior detection (fighting, falling, running, crouching), and frictionless access control. This technology is engineered to perform reliably even in challenging conditions such as poor lighting, difficult camera angles, and crowded environments, effectively minimizing false positives.

The system supports integration with existing camera and VMS systems and offers flexible deployment options, including cloud, on-premises via the Vision AI Appliance, or through an SDK. Oosto emphasizes ethical AI development, focusing on reducing bias through extensive training on diverse datasets and incorporating privacy protection measures like data encryption and blurring of non-watchlisted individuals. It aims to empower organizations across various sectors to proactively address security threats, enhance operational safety, and improve efficiency.

Features

Accurate Person of Interest Alerting: Identifies persons of interest in real-time, even in challenging conditions ("in-the-wild").
Smart Behavior Detection: Detects complex actions like fighting, falling, running, walking, crouching, lying down, line-crossing, and pattern anomalies.
Edge-to-Cloud Deployment: Offers flexible deployment via cloud, on-premises (Vision AI Appliance), or SDK.
Frictionless Access Control: Uses existing cameras for touchless entry authorization (OnAccess).
Real-Time Watchlist Alerting: Provides live alerts based on custom watchlists (OnWatch).
Body and Skeleton Analysis: Combines body and head data for detailed person detection and re-identification.
Pose and Action Detection: Contextual action recognition (fighting, falling, running, etc.).
Liveness Detection: Detects spoof attempts using passive 2D/active 3D liveness for remote authentication.
Privacy Protection: Features include blurring non-watchlisted individuals and data encryption (image-to-vector conversion).
Bias Mitigation: Algorithms trained on diverse datasets to reduce demographic bias.

Use Cases

Enhancing public safety in airports, seaports, smart cities, and correctional facilities.
Improving security in casinos, corporate buildings, retail stores, and stadiums.
Securing critical infrastructure, manufacturing plants, and healthcare facilities.
Implementing touchless access control systems.
Conducting forensic video analysis.
Reducing false security alarms.
Preventing retail loss.
Monitoring operational safety and perimeter security.
Tracking persons of interest across multiple cameras.
Managing time and attendance.
Implementing advanced visitor management systems.