3DiVi Releases Omni Agent 3.0 with Custom Video Analytics Scenarios, New Detection Objects and Hardware Cost Savings
3DiVi, an international provider of AI-powered computer vision solutions, today announced the release of Omni Agent 3.0, a major update to its video analytics platform that makes detection more adaptive, accurate, and cost-efficient than ever. This release introduces a redesigned architecture, optimized server performance, enhanced low-power device support, and a fully revamped web interface for streamlined configuration and monitoring.
Covina, CA, November 20, 2025 --(PR.com)-- New Architecture & Smarter Detection Capabilities
Omni Agent 3.0 features a new pipeline management architecture that allows users to create custom video analytics scenarios using existing detectors—without the need to reinstall the agent. Detection scenarios have been expanded to include mobile phones, specialized objects, and unusual activities (UAD), with seamless updates to detection weights and templates.
Users can now optimize detection scenarios for specific camera angles—overview or vertical—directly via the web interface. Detector accuracy and coverage have been improved, while resource consumption has been reduced for body, skeleton, and human activity recognition (HAR) tracking. Retrained models now allow recognition of people and movements in challenging lighting conditions, both indoors and outdoors. These upgrades are already proving valuable in industrial projects, where perimeter control and interaction with equipment directly impact safety and operational efficiency.
Additionally, Omni Agent 3.0 allows users to define object sizes for ROI and line-crossing triggers, and set timeouts for repeated events, reducing false alarms and operator workload.
Hardware Cost Savings & Improved Low-Power Performance
Analytics scenarios can now be applied to selected cameras only, with customizable settings and event triggers per camera. Users can restrict analysis to specific zones, define ROIs within those zones, and mask irrelevant areas while retaining full-frame recordings for triggered events. An external API is now available to start or stop selected analytics scenarios.
The update also brings support for TensorRT 8x acceleration on Linux, while Windows continues to support CUDA acceleration, further improving performance on low-power hardware.
Web Interface Updates – Easier Configuration & Monitoring
All configuration settings have been migrated from files to the web interface, making it faster and simpler to create and manage custom analytics scenarios. The interface now displays workspace and licensing information, allows toggling object rendering on previews, and provides real-time monitoring of analytics scenario loads.
Omni Agent 3.0 empowers businesses to make faster, smarter decisions with video analytics while reducing hardware costs and operational workload. For more information or to schedule a demo, please contact the 3DiVi team: https://3divi.ai/products/omni-agent
Omni Agent 3.0 features a new pipeline management architecture that allows users to create custom video analytics scenarios using existing detectors—without the need to reinstall the agent. Detection scenarios have been expanded to include mobile phones, specialized objects, and unusual activities (UAD), with seamless updates to detection weights and templates.
Users can now optimize detection scenarios for specific camera angles—overview or vertical—directly via the web interface. Detector accuracy and coverage have been improved, while resource consumption has been reduced for body, skeleton, and human activity recognition (HAR) tracking. Retrained models now allow recognition of people and movements in challenging lighting conditions, both indoors and outdoors. These upgrades are already proving valuable in industrial projects, where perimeter control and interaction with equipment directly impact safety and operational efficiency.
Additionally, Omni Agent 3.0 allows users to define object sizes for ROI and line-crossing triggers, and set timeouts for repeated events, reducing false alarms and operator workload.
Hardware Cost Savings & Improved Low-Power Performance
Analytics scenarios can now be applied to selected cameras only, with customizable settings and event triggers per camera. Users can restrict analysis to specific zones, define ROIs within those zones, and mask irrelevant areas while retaining full-frame recordings for triggered events. An external API is now available to start or stop selected analytics scenarios.
The update also brings support for TensorRT 8x acceleration on Linux, while Windows continues to support CUDA acceleration, further improving performance on low-power hardware.
Web Interface Updates – Easier Configuration & Monitoring
All configuration settings have been migrated from files to the web interface, making it faster and simpler to create and manage custom analytics scenarios. The interface now displays workspace and licensing information, allows toggling object rendering on previews, and provides real-time monitoring of analytics scenario loads.
Omni Agent 3.0 empowers businesses to make faster, smarter decisions with video analytics while reducing hardware costs and operational workload. For more information or to schedule a demo, please contact the 3DiVi team: https://3divi.ai/products/omni-agent
Contact
3DiVi Inc.
David Mitchell
+7 (351) 211-5099
3divi.ai
David Mitchell
+7 (351) 211-5099
3divi.ai
Contact
Categories