AI solutions help reduce traffic accidents
In this article (5)
VinAI’s Driver Technology Software provides drivers with a 360-degree view of all surroundings, including blind spots, to help avoid traffic accidents.
At the recent 2023 Consumer Electronics Show (CES) in Las Vegas (U.S.), VinAI showcased two smart driving features – Automatic Mirror Adjustment (AMA) and Jelly View, a 360-degree view mode. AMA is the world’s first auto mirror adjustment feature, according to a VinAI announcement.
These are the two newest features of the Driver and Occupants Monitoring System (DMS/OMS) and Advanced Surround View Monitoring (SVM). Both are in-vehicle solution suites that apply AI to reduce traffic accidents and improve the driving experience.
VinAI applies infrared cameras, vehicle sensors, and facial expression analysis technology (ranked 6th in the world), which detect the driver’s condition, such as drowsiness or fatigue, and also monitors for risky behaviors, including being distracted or using a phone. As soon as the system detects such issues, the DMS issues different levels of warnings throughout the journey, even when the driver is wearing sunglasses or a mask.DMS is compatible with various hardware platforms, enabling carmakers to easily integrate it into their existing systems without spending too much on transformation.
The Automatic Mirror Adjustment (AMA) is also an outstanding feature, easily activated with just the touch of a button. AMA can automatically adjust the mirrors on both sides of the car to the perfect position, thanks to the driver’s eye position analysis.
The system works by estimating the driver’s eye position three-dimensionally. Normally, this problem would require two or more cameras to solve, but VinAI’s advanced system can accurately predict eye position with a single infrared camera.
The VinAI enhanced 360-degree panoramic view feature (ASVM) helps the driver keep a close eye on the outside world while driving. The driver can also adjust the view using the on-screen control panel by choosing Jelly View mode, a transparent wraparound view. With the support of Jelly View, the driver can observe the entire space surrounding the vehicle: the front, rear, behind and outside, from the trunk to the chassis.
The system uses AI technology to collect images from four fisheye cameras as well as information from the accelerometer to construct the vehicle’s chassis view in real-time. This helps drivers observe blind spots through the rearview mirrors. From there, users can easily detect obstacles and avoid unexpected accidents on narrow roads, rough roads, or in parking lots.
“The VinAI enhanced 360-degree panoramic view feature (ASVM) helps the driver keep a close eye on the outside world while driving.”
The SVM system won the VnExpress Innovation Awards 2022 and is currently being deployed in the VFe34 electric car in Vietnam. The Jelly View feature was first introduced to the public at CES 2023.
Dr. Bui Hai Hung, General Director of VinAI Artificial Intelligence Research and Application Joint Stock Company, said the new AMA and Jelly View features had received positive reviews from professionals as well as consumers at CES 2023.
These AI solutions can significantly reduce accidents, which are a big issue not just for Vietnam, but also for the world, including developed countries. This “life-saving” technology is highly applicable.
VinAI has also developed various AI products in many other fields including smart cities. One such solution includes Smart Edge, an access control solution that uses facial recognition technology. Smart Edge is currently deployed to more than 50,000 people at Vinpearl resorts, Vinhomes Ocean Park, Vinhomes Smart City, and Vinhomes Grand Park. The feature has proven to be over 99% accurate.
Just three years since its establishment, VinAI has registered 66 patents in Vietnam and the U.S. and has published 107 research papers at the world’s leading AI conferences.
VinAI was the only Vietnamese company ranked by Thundermark Capital as one of last year’s top 20 leading global AI research companies.
VinAI’s cutting-edge machine learning-based technology aims to address global needs in various industries, with a focus on automobiles and smart cities.
VinAI’s cutting-edge machine learning-based technology aims to address global needs in various industries, with a focus on automobiles and smart cities.
VinAI’s cutting-edge driving technologies, including AMA and Jelly View, were showcased at CES 2023 from January 5 to January 8, offering guests the opportunity to experience these features first-hand in two electric cars, VinFast’s VF8 and VF9. The vehicles were installed with the full set of driver monitoring systems and 360-degree panoramic view systems.
Questions & Answers
Q.Which specific VinAI products help reduce traffic accidents?
Which specific VinAI products help reduce traffic accidents?
VinAI's Driver Technology Software, including the Driver and Occupants Monitoring System (DMS/OMS) and Advanced Surround View Monitoring (SVM) with features like AMA and Jelly View, helps to reduce traffic accidents.
Q.How does the Automatic Mirror Adjustment (AMA) feature work?
How does the Automatic Mirror Adjustment (AMA) feature work?
AMA automatically adjusts the car's side mirrors to the perfect position by analysing the driver's eye position. It uses a single infrared camera to accurately predict this three-dimensionally.
Q.What is Jelly View and how does it assist drivers?
What is Jelly View and how does it assist drivers?
Jelly View is a transparent wraparound view that allows drivers to observe the entire space around the vehicle, including blind spots. It collects images from four fisheye cameras and accelerometer data.
Q.Are these new driving technologies currently in use in any vehicles?
Are these new driving technologies currently in use in any vehicles?
Yes, the SVM system is currently being deployed in the VFe34 electric car in Vietnam. The AMA and Jelly View features were showed at CES 2023 in VinFast's VF8 and VF9 electric cars.
Reader pulse
Is AI in vehicle safety a major shift?
19,567 votes so far