StradVision to collaborate with leading custom SoC supplier Socionext to provide efficient ADAS solutions
The respective pioneers in AI-based camera perception software and System-on-Chip design and development will work together to bring state-of-the-art object recognition to the global automotive industry
StradVision has announced a collaboration with Socionext Inc., a leading supplier of System-on-Chip (SoC) solutions, to bring StradVision’s deep learning-based camera perception software SVNet, to the global market.
Through this collaboration, the two companies will provide state-of-the-art object recognition technology with software and hardware specialized for deep learning to the Advanced Driver Assistance Systems (ADAS) and autonomous driving markets.
“This collaboration allows us to optimize the integration of our perception software into SoC developed by Socionext, resulting in extremely stable software operation within vehicles. SVNet running on Socionext’s SoC will be a great solution for the ADAS and autonomous markets where robust performance and safety are essential,” said Junhwan Kim, CEO of StradVision.
StradVision’s pioneering SVNet software allows vehicles to detect and recognize objects on the roads. It performs well even in harsh weather conditions and is able to prevent traffic accidents by processing collected road data with high speed and accuracy.
Compared with competitors, SVNet is compact, requires dramatically less memory capacity to run, and consumes less electricity. It can also be customised for any hardware system thanks to StradVision’s patented and cutting-edge Deep Neural Network enabled software.
SVNet is currently used in mass production models of ADAS and autonomous driving vehicles that support safety function Levels 2 to 4, and will be deployed in more than 8.8 million vehicles worldwide.
“Socionext has a proven track record of supplying a large number of custom SoCs for various in-vehicle applications, offering cutting-edge design development technology including support for miniaturization processes and quality control,” said Koichi Yamashita, Head of the Automotive Business Unit, Automotive & Industrial Business Group at Socionext.
“It is now possible to provide object detection with high recognition accuracy that is optimized for vehicle-mounted camera systems. Software like SVNet can detect vehicles, pedestrians, lanes, and free space. In the automotive market, where significant growth is expected in the future, we will actively utilize StradVision’s software with the aim of providing custom SoCs that are a source of customer differentiation and competitiveness.”