Quality and Performance at one place

AnyVision utilizes NVidia hardware to achieve high-speed, real-time face recognition from surveillance video streams.  Our system is highly optimized for GPU acceleration allowing us to deliver real-time analysis of streaming data whilst achieving unprecedented accuracy.

Video Optimized

AnyVision uses proprietary algorithms to achieve world-class performance on video data. Be it streaming surveillance or forensic video analysis, NVidia GPUs permit real-time analysis enabling pro-active threat intervention.

Scalable Architecture

Process multiple high-resolution cameras on a single system. Using NVidia GPUs a single AnyVision server can currently process in excess of 40 video streams (1920×1080 pixels) concurrently and in real-time, and soon to reach over 150+.


Although designed for Video, AnyVision’s accuracy remains world-leading even for single-frame analysis. This means that installations can be configured to maximize performance for different conditions including camera resolution and FPS, density of faces, and field of view. Strong performance can even be achieved on streams with low FPS meaning AnyVision’s solution can be utilized across a broad range of existing installations.

Example benchmarks are detailed below with consumer-grade NVidia hardware. Server-class hardware such
as the Tesla series allows very high numbers of concurrent streams to be processed.

Database size Resolution Processing Time
50K faces 1080p 11ms
500K faces 1080p 31ms
1M faces 1080p 50ms
3M faces 1080p 118ms
8M faces 1080p 270ms

Nvidia Metropolis™

Anyvision is an official partner with Nvidia on the Metropolis intelligent video analytics platform.

Nvidia Metropolis™ is bringing advanced AI to cities around the world. Anyvision is proud to be selected as the facial recognition engine to support this exciting initiative.

Related Posts