Human-centric
AI Computer Vision
Protect access to your workplace with FR
3D body (skeletal) tracking middleware
Edge AI hardware
for face and body tracking
AI video analytics platform
for human activity recognition
Advanced face liveness detection for digital onboarding and KYC
BAF
Scalable API for facial
and body recognition
Facial and body recognition library for server, mobile and embedded solutions
3DiVi Inc., founded in 2011, is one of the leading developers of AI and machine learning (ML) technologies for computer vision.
3DiVi Blog

Even Better Face Recognition & Human Activity Tracking for Safe City & Manufacturing Safety

The latest update of 3DiVi Inc.'s OMNI Platform brings remarkable enhancements particularly beneficial for such projects as 'Safe City' or 'Safe Manufacturing Facility'.

Enhanced Event Service Functionality

The OMNI Platform now includes expanded functionalities in its event service. This is a game changer for safety and security projects in urban and manufacturing settings. By integrating new types of events (HAR, ROI, Crossing), the platform ensures more comprehensive monitoring and analysis capabilities.

Web Interface Improvements

The addition of event and WebSocket pages in the web interface is a significant step forward. It reduces the time and effort required by clients to understand and test the product's functionalities. This user-friendly approach is crucial for fast-paced urban and industrial environments where efficiency and ease of use are key.

GraphiQL API Access and WebSocket Enhancements

Providing a direct link to the GraphiQL API of the event service streamlines the process of integrating and utilizing the platform in complex security networks. The WebSocket has become a unified point for receiving all types of detections and identifications, which simplifies the data acquisition process.

Advanced Filtering and Identification

The ability to filter by profile ID and the addition of an 'activityId' in the Event Service API enhances the platform's capability to link events to their originating activities. This feature allows for a more detailed analysis, including tracking duration, attention metrics, demographic information, and emotional states.

Customizable Event Size and Storage Solutions

The introduction of the 'event_max_body_size' parameter addresses the challenges posed by high-resolution cameras in large-scale projects. This customization ensures that full frames are efficiently processed, catering to projects with extensive camera networks. Furthermore, the integration of MinIO object storage for event images significantly boosts the efficiency of recording results from numerous cameras, while also providing a scalable solution for storing heavy data separate from the main database.

Even Better Data Protection and Security

A major enhancement is the implementation of endpoint_id for WebSocket authentication, replacing the conventional platform access token method. This shift not only streamlines the authentication process but also bolsters security by reducing the exposure of sensitive access credentials.
Furthermore OMNI Platform now allows the assignment of specific IDs during camera setup, thereby eliminating the need to disclose RTSP stream credentials. This enhances privacy in surveillance systems.
The update also includes a token access reset function in the Django admin panel, crucial for maintaining system integrity, especially in scenarios of administrative change or token compromise.

Enhanced Stability

The principle of activity and event data cleansing has been changed to improve stability during long-term operation of the platform.

Better User Experience

The new version introduces significant web interface improvements to enhance user experience and efficiency. Key updates include enriched system status indicators on the main dashboard, a revamped navigation panel, and refined content filtering elements.
Notably, the platform now features an image-based search for profiles and activities, enabling more effective retrospective searches. Additionally, string value filters have been optimized for quicker, real-time search results, streamlining the overall user interaction with the platform. These enhancements are designed to reduce the time and effort clients spend on understanding and testing the product, emphasizing 3DiVi Inc.'s commitment to user-centric design in computer vision and security software.

Try OMNI Platform for free >