®
KINEVA® / FOR TECHNICAL TEAMS
Turn what happens in an image into a finding.
REBOTNIX® trains visual models for specific detection tasks: objects, damage and surfaces. They deliver the first results that HORIZON and KINEVA® One can work with.
← Back to KINEVAKINEVA® / Visual models / CNN
How this component works.
Capture imagery
Cameras observe the relevant part of a city or production process.
Run the model locally
GUSTAV processes the image with the appropriate REBOTNIX® model on site.
Pass on the result
HORIZON receives the finding; KINEVA® One can interpret the combined picture from many findings.
KINEVA® / THERMAL
Detect when it is dark.
Thermal cameras provide temperature imagery without visible light. The video shows how REBOTNIX® models analyse this data on site and reveal thermal anomalies.
Explore thermal monitoringKINEVA® / COMPUTER VISION
What the camera sees in the city.
This example shows visual segmentation of a street scene: the model distinguishes the road, vehicles, buildings, trees and other objects. These results provide a foundation for specific HORIZON Smart City tasks.
Explore HORIZON Smart CityREBOTNIX® / KINEVA®
Part of a maintained model family.
REBOTNIX® develops and maintains its models and tools. GUSTAV brings processing on site, HORIZON connects results to workflows, and KINEVA® One answers questions about the combined data.
10-year lifecycle support for fine-tuning and optimization of our model families.
KINEVA® One ↗KINEVA®
A technical question?
Talk to our team about models, integration and the right place to run them.
Contact the technical team