FPGA-based Intelligent Video Analytics (IVA)
Global AI solution customer specializes in facial recognition technology in video surveillance applications for the security and public safety sector. The customer requires the performance and parallel processing capability to not only recognize moving images but also efficiently translate unstructured real-time video data into contextual and analytical insight.
Inspur’s F10A FPGA accelerator was used to boost the deep learning capabilities of the customer’s platform. The result is an FPGA-based intelligent video analytics (IVA) solution with high density, high performance and high reliability. It demonstrated the video processing performance of 1080P at 25fps over 16 channels.
The solution gives the company the performance, speed and inferencing power it needs to capture and analyze real-time footage, furthering its ability to develop comprehensive video security, intelligence, and investigative capabilities for security and public safety organizations.