VIA Technologies is partnering with AI vision startup Lucid, to deliver AI-based depth sensing capabilities to more dual- and multi-camera devices in the security, retail, robotics and autonomous vehicle space.

With Lucid’s proprietary 3D Fusion Technology embedded into the VIA Edge AI 3D Developer Kit, security and retail cameras, robots, drones, and autonomous vehicles will be able to capture accurate depth and 3D with dual- or multi-camera setups while reducing the costs, power, and space consumption of previous hardware depth solutions. As VIA looks to build out its Edge AI solutions roadmap, Lucid will be adding camera- and machine-learning based depth capabilities on top of every platform.

The AI-enhanced 3D/depth solution developed by Lucid, known as 3D Fusion Technology, is currently deployed in devices such as 3D cameras, security cameras, robots, and mobile phones. In the VIA Edge AI 3D Developer Kit, the AI depth solution runs on the Qualcomm APQ8096SG embedded processor, which features the Qualcomm AI Engine along with support for multiple cameras to help Lucid provide superior performance compared to other hardware depth solutions and deliver an industry-leading and unique pure machine learning-based software solution.

“Through this exciting partnership, VIA and Lucid together are tackling the fast scaling required by depth sensing with our Edge AI platforms and systems. We see major applications not only for security cameras, but also smart retail, robots, drones, and especially autonomous vehicles,” said Richard Brown, Vice-President of International Marketing, VIA Technologies. “This collaboration will yield a hardware/software combination where VIA provides customisable and embeddable Edge AI developer kits, systems, and cameras with Lucid’s AI solution as an integrated software module. This will provide a tightly-integrated package for all device makers looking to kickstart the development of depth sensing capabilities.”

In the current market, smart cameras with depth sensing capabilities have become critical for applications in the security, retail, robotics and IOT industries, with applications such as more refined facial identification accuracy to determine distances between people in the space beyond glass walls and precise object recognition, tracking and measurements.

For robots, drones and autonomous vehicles, depth sensing has dramatically improved the capability to navigate, save and reconstruct spaces in 3D, providing more autonomy for machines.

The current hardware solutions, emission- and laser-based, impact the total cost, size and power consumption of devices in particular. The VIA and LUCID solution will not only help eliminate these challenges but also help to speed up system scaling and deployment by enabling the integration of depth sensing capabilities within a much lower cost, form factor, and power budget.