Paravision enhances video facial recognition with Streaming Container update

Paravision has developed a new version of its Streaming Container technology to deliver its facial recognition engine with enhanced efficiency and accuracy within standard IP video feeds.

The new Docker container-based system is directly compatible with standard video protocols like RTSP and capable of enabling multi-factor authentication or biometric identification for access control or complementary video security applications, the company says.

“The desire for touchless, frictionless access is now finally matched by solutions which are accurate, robust, fast, and cost-effective,” says Paravision President and COO Benji Hutchison in an interview with Biometric Update on the Converge 2021 announcements. “Many of these products are currently in development with our partners, so expect some exciting things in the year to come.”

The Paravision Streaming Container is also fully modular and is designed to be deployed alongside Paravision’s other cloud-ready, Docker container-based products.

The new solution can therefore be deployed either to enable complete solutions performing facial recognition from video, or in a standalone manner to extract face images from video feeds.

Technically, the Streaming Container is based on Paravision’s facial recognition algorithms, which stood out in recent testing by NIST for matching performance across the full range of FRVT biometric verification categories.

The latest version of the Streaming Container released by Paravision is version 5, which provides several improvements over previous versions.

These include the handling of over 250 frames per second of video, coming from any number of sources, and on a single Intel CPU or NVIDIA GPU instance.

In addition, Streaming now offers advanced face tracking, which the company says dramatically improves the ability to continue following faces through a video stream even if they are momentarily blocked or leave the frame. This reduces repeat identification for more consistent results with a lower burden on the backend matcher.

Paravision Streaming Container 5 can be either deployed on-premises or via the cloud and supports both Linux and Windows operating systems.

The solution also supports Intel CPU (OpenVINO) and  NVIDIA GPU (TensorRT) computer vision frameworks, as well as RTSP and HTTP video protocols.

The release of the new software comes weeks after Paravision added Kurt Takahashi to its corporate board of directors.

Article Topics

access control  |  authentication  |  biometric identification  |  biometrics  |  facial recognition  |  Paravision  |  video surveillance