4Sense Dragonfly

The first AI system for recognizing human behavior

The Dragonfly System

4Sense has made a giant leap in computer vision where seeing and learning in 4D is done through multi-spectrum sensing and advanced machine learning. The Dragonfly system mimics the dragonfly's advanced visual anatomy with three components: 4D Sensors emulate its eyes, a Mesh Network parallels its fast-reflex neural system, and a Cortex Cloud corresponds to its brain.

4D Sensors

Small, low-power sensors use advanced machine learning to enable multi-spectrum computer vision. Spectral Fusion technology combines the strengths of RGB, solid-state LIDAR, thermal and audio to track people with centimeter-level precision. And Deep Edge Processing performs all image processing and machine learning within each sensor, allowing for only digital metadata output.

Mesh Network

Mesh Intelligence™ unifies the machine learning from all 4D Sensors to enable system-wide learning across the mesh network. The sensor system provides intelligence for inter-sensor tuning, floor plan calibration and continuous person recognition and tracking.

Cortex Cloud

A motion analytics engine provides path, pattern and behavior data for real-time and post-event business intelligence. Body Prints™ help to differentiate people like a fingerprint, and include body position (standing, seated, running, fallen), hair and clothing color and styles, height, gait, voice and 3rd party authentication information.

System Flexibility

Dragonfly requires only 4D Sensors to be installed on-premises for new, retro-fit, and temporary deployments. Cloud APIs provide easy access to rich 4D data, and enable custom solutions through integrations with identity, security, automation, and point-of-sale (POS) systems. 

Working with 4Sense

If you're a customer interested in learning about our early adopter program, or if you're a partner and would like to work with us or license our technology, please contact us.