Hailo, Lanner pair on edge systems for rapid face biometrics processing

Hailo Technologies Ltd, a developer of specialized deep learning processors, has partnered with Lanner Electronics, a manufacturing and design firm, to build edge gateway devices capable of processing high volumes of video streams at the network edge. Facial recognition in smart city and retail environments are one potential beneficiary of the edge computing systems the pair plan to offer.

The amount of data coming from multiple video feeds in a surveillance system can overwhelm the internet connections used to send data to the central cloud for processing. Hailo’s chip enables high performance AI capabilities such as running face biometrics algorithms of 26 Tera Operations Per Second (TOPS) in edge devices while consuming less power than other chips.

The Hailo-8 AI module chip will be incorporated into Lanner’s LEC-2290 edge box, which is based on an x86 processor. The chip will also be offered in Lanner’s small-footprint LEC-7242 industrial wireless gateway integrates the Hailo-8 into a fanless appliance for real-time edge AI applications. Both solutions process multiple video streams in real-time on a single device, while also enabling secure network connectivity.

Last year, Hailo closed a $60 million round of venture funding for its edge AI processor amidst a hot market for GPUs and other chips that can be used to securely accelerate applications such as facial recognition. Market research company Omdia forecasts that global AI edge chipset revenue will grow from $7.7 billion in 2019 to $51.9 billion by 2025, and edge facial recognition has been forecast to generate $2.3 billion a year by the then. One gating factor for Hailo and Lanner: a global semiconductor shortage that CEOs from Nvidia and Intel say could last for another year or two.

Article Topics

AI chips  |  biometrics  |  biometrics at the edge  |  deep learning  |  edge computing  |  facial recognition  |  Hailo  |  Lanner Electronics  |  retail biometrics  |  smart cities  |  video surveillance