Built-in Analytics Modules With Advanced Deep Learning Algorithms
Monitor Onsite Occupancy
Upgrade onsite security by identifying and authenticating specific people
Real-Time Facial Recognition
Accurately identify faces from the face database using a powerful deep-learning algorithm with evolving adaptive precision, proven to achieve > 97% accuracy2 in real-world settings.
Instantly identify and record multiple faces simultaneously3.
Authentication with Customizable Face Database
Validate your detection accuracy
With Parameter Adjuster, you can simulate actual scenarios using previous recordings or detection results. This ensures that the parameter settings are validated and optimized in the installation phase to achieve the best accuracy. Learn more
Server-based video analytics
Compared to cloud-based or camera-based models, running video analytics on Synology DVA3219 provides several distinct advantages:
Security & privacy
No footage is uploaded to the cloud for analysis purposes, removing concerns over sensitive information and regulatory requirements.
No recurring fees
Just a one-time hardware purchase – no monthly subscriptions or additional charges for future software updates.
Instead of post-event analysis, DVA3219 notifies security operators as an incident takes place, allowing them to take immediate actions. No need to worry about unstable Internet connections or bandwidth usage, either.
Ease of integration
Compatible with over 7,500 IP cameras, there’s no need to replace them to gain analytics functionality – simply substitute DVA3219 for your existing NVR.
- Features described here are available on DVA series models only.
- Testing is done by the National Institute of Standards and Technology (NIST) in their Facial Recognition Vendor Test (FRVT) under its WILD category (using faces extracted from real-world footage).
- Identify and record up to 25 faces per frame per camera channel and up to 4 channels for each DVA model running Surveillance Station. (Live Viewing via VS360HD does not support displaying more than 25 facial recognition results simultaneously.)