Computer Vision Developer StradVision to Showcase its Most Advanced Perception Camera for Autonomous Driving & ADAS at Auto Tech 2021
【Summary】StradVision, a pioneer in AI-based vision processing technology for autonomous vehicles and automotive ADAS systems, will showcase the latest features of its camera perception software named “SVNet” during The China Guangzhou International Automotive Technology Expo 2021, which is being held May 25-27.
StradVision, a pioneer in AI-based vision processing technology for autonomous vehicles and automotive ADAS systems, will showcase the latest features of its camera perception software named "SVNet" during The China Guangzhou International Automotive Technology Expo 2021, which is being held May 25 -27.
SVNet deep learning-based perception software that supports advanced driver assist systems (ADAS) features such as Highway Driving Assist and Automated Valet Parking.
This year's Auto Tech will be held at Guangzhou Poly World Trade Center Exhibition Hall in China. The annual expo covers all the important topics of the automotive industry such as electric vehicles, automotive electronics, connected cars, autonomous driving technology and more.
At Auto Tech 2021 next week, StradVision will demonstrate its front-facing camera solutions based on the NVIDIA Xavier system on a ship (SoC). In addition, it will expand its reach to the latest technologies such as Surround View Monitoring and "Pseudo Lidar", which is attracting a great attention in the industry.
In 2019, Cornell University researchers developed a method of using low-cost, stereo cameras that allow autonomous vehicles to detect 3D objects with a range and accuracy approaching that of lidar. It essentially treats camera output as if it were lidar data.
Pseudo lidar technology enables depth estimation of objects using cameras without expensive and high-performance lidar equipment. The technology mimics how traditional lidar scans an environment. It has the potential to significantly reduce the use of expensive lidar hardware for building self-driving capable vehicles at scale.
StradVision improved the perception capability and range of its AI-based camera perception systems by adding higher resolution images of 8 megapixels (3840x2160), as well as the ground-breaking Highway Driving Assist (HDA) feature developed by automaker Hyundai that's based on this.
The new and improved multi-camera 360-degree perception system uses up to nine cameras, to support Level 3 and above autonomous driving features.
HDA is available on select Hyundai, Kia and Genesis vehicles and supports automated highway driving. It reduces driver burden on highways by controlling the vehicle's steering, acceleration and deceleration to maintain the speed set by the driver, keep a safe distance from the vehicle directly ahead, and keep the vehicle in the center of its current lane.
StradVision's computer vision and AR technology is supported by its SVNet deep-learning-based perception software, which supports enhanced depth-mapping solutions, semantic segmentation, and lane detection for lane-keep assist and other vehicle ADAS safety features.
SVNet, a deep learning-based perception software. It enables vehicles to better execute self-driving functions by improving perception capabilities of vehicle cameras. The software relies on deep learning-based perception algorithms which use less compute resources and power, both of which are concerns for automakers.
SVNet allows vehicles to detect and identify objects accurately from camera images, such as other vehicles, lanes, pedestrians, buildings, free space, road signs, and traffic lights. It works even in harsh weather conditions or poor lighting.
StradVision's SVNet supports more than 14 hardware platforms, including processors from Renesas, NVIDIA, Qualcomm, Xilinx and Texas Instruments. It can also be customized and optimized for any other hardware system using StradVision's patented Deep Neural Network-enabled technology.
SVNet is currently used in mass production models equipped with ADAS and autonomous driving features. It can support driving SAE automated driving levels 2 to 4, the company said.
Nvidia's DRIVE Xavier SoC used by Stravision is widely used for autonomous vehicle development. In addition to providing the robust processing hardware, Nvidia provides its DRIVE customers with pre-trained AI models and training code.
For developers of self-driving vehicles, these AI models include traffic signal recognition, pedestrian detection, path perception, and gaze detection, which is used for driver monitoring systems. Developers can also customize these models to increase the perception capabilities of their own self-driving systems.
StradVision will supply its augmented reality (AR) technology for Navigation and Lane Keeping Assistance Systems (LKAS). The second partnership is with a leading global Tier-1 supplier to provide a Surround View Monitoring (SVM) algorithm that supports Park-Assist functions such as Automatic Parking Assistance (APA), which are being introduced on many new vehicle models today.
StradVision declined to name the companies at this stage, but said its AR technology will be featured on more than 40 vehicles from the OEM, the company said. The vehicles are scheduled for production globally in 2022.
StradVision was founded in 2014. The company is an automotive industry pioneer in AI-based vision processing technology. The company says its accelerating the deployment of fully autonomous vehicles by making robust ADAS features available at a fraction of the market cost compared with competitors.
The company is based in Seoul, South Korea, with additional offices in Silicon Valley, Tokyo, and Munich.
StradVision recently received the 2020 Autonomous Vehicle Technology ACES Award in Autonomy in the software category. The ACES Awards recognize innovative technologies in the industry, such as autonomous driving and vehicle connectivity.
Originally hailing from New Jersey, Eric is a automotive & technology reporter covering the high-tech industry here in Silicon Valley. He has over 15 years of automotive experience and a bachelors degree in computer science. These skills, combined with technical writing and news reporting, allows him to fully understand and identify new and innovative technologies in the auto industry and beyond. He has worked at Uber on self-driving cars and as a technical writer, helping people to understand and work with technology.
Baidu Inc. to Deploy 1,000 ‘Apollo Moon’ Level-4 Robotaxis Over the Next 3 Years in China
Ford Acquires California Startup Electriphi to Manage Charging for its Commercial Electric Vehicle Customers
Alphabet’s Autonomous Driving Unit Waymo Announces a New $2.5 Billion Funding Round
Continental & Elektrobit Announce the First In-Vehicle Integration of Amazon’s ‘Alexa Custom Assistant’
Ford Motor Co Officially Starts Production of the Rugged New Bronco SUV
GM’s Autonomous Driving Unit Cruise to Tap $5 Billion Credit Line to Mass Produce the Origin Autonomous Shuttle
Lidar Pioneer Velodyne Unveils its Next-Generation Velabit Sensor
CEO, CFO of Electric Truck Startup Lordstown Motors Abruptly Resign
- Some Ford Mustang Mach-E Owners Encountering Battery Bug
- BMW Shares New Details & Pricing for the Electric i4 Sedan, Will Start at $55,400 in the U.S.
- TomTom Announces ‘Navigation for Automotive’ a Cloud-based Solution for Automakers
- Future Volkswagen Electric Models Will Get Bidirectional Charging Capability
- Semiconductor Shortage Could Result in 1.3 Million Fewer Vehicles in 2021
- China’s Baidu Announces ‘Apollo Air’ Which Supports Self-driving Vehicles Using Real-time Data From 5G-Connected Roadside Units
- Sweden-based EV Battery Startup Northvolt, Founded by Former Tesla Execs, Raises $2.75 Billion in New Equity
- Japan’s Honda Commits to Selling Only Electric Vehicles by 2040
- China’s Baidu Inc. Plans to Have its Apollo Autonomous Driving Technology Pre-installed in One Million Vehicles in 3 to 5 Years
- Ford Reveals the Maverick Pickup With Standard Hybrid Powertrain Starting at Just $19,995