See a few cool things at the GPU Technology Convention
【Summary】See a few cool things at the GPU Technology Convention.
As one the largest and most important events of the year for software developers and industry professionals, pioneering technology and progress in fields such as augmented reality, autonomous driving, deep learning, and artifical intelligence were on display at the 2017 GPU Technology Convention. Futurecar reporters are onsite to bring news related to the auto industry.
AI Police Cars of the Future
Suppose a future when police are equipped with a more intelligent car to patrol and maintain the city's order. Law enforcement will have greater insights provided by surveillance data gathered by the car's sensors to detect any suspicious activity. The real-time pedestrian and vehicle activity tracking system in this car can scan your license plate to see if it's expired or not, or detect a suspicious person lurking nearby. The AI police will be a powerful tool for combating crime and keeping the public safe.
Inside the car, the display scans the vicinity to track activity and identify people using facial recognition software.
On the roof of the vehicle, a drone is waiting to be dispatched on occasions when remote surveillance is needed.
Autonomous driving truck
Nvidia has partnered with Paccar to develop a self-driving truck. The brain of the vehicle is powered by the Nvidia Drive PX 2 system, a compact in-vehicle supercomputer built for self-driving. It's powerful enough to process inputs from up to 12 video cameras, lidar, radar, and ultrasonic sensors, which means, the the Nvidia Drive PX 2 can control lane changes, speed, object detection, and adaptive cruise control to slow down and avoid collisions.
Shopping mall patrol robots
AI robots in the future may gather and process real world data inputs and make decisions instantly facilitated by deep learning algorithms.
This Knightscope autonomous security robot can monitor crime in various locations by running real-time video analytics powered by Nvidia Jetson. The robot is equipped with a variety of 360 degree sensors, including a range finder, a camera to detect people and compare them with an active criminal list, or to aid in routine city patrols.
Nvidia AI Co-Pilot Car
Four features are included in Nvidia's AI Co-Pilot system, including facial recognition, head tracking, gaze tracking, and lip-reading.
"We're assisting the driver in having a better driving experience. I identify you as the driver, I can give you a much better experience. I will know what your favorite radio station is, I may know at this time every day you drive to work, so I can navigate you there." Elizabeth Sanville from Nvidia's marketing team explained to a Futurecar reporter.
She added that enabling gaze and head tracking is observing the driver's driving habits, and may be able alert the driver to prevent an accident from happening, e.g, missing a red light, failing to see a cyclist on your right, etc.
The most amazing feature of the Co-Pilot system is that it can read your lips.
According to Nvidia, the lip reading software combined with speech recognition can now reach 95% accuracy.
"Speech recognition works really well in a quiet room, but it doesn't work that well in a car where the music may be on, people may be speaking, kids are playing in back…so with lip reading focusing on the driver, it can really increase the accuracy." Sanville said.
Safer Trident: the semi autonomous vehicle for divers
Using AI for underwater navigation and coral detection, the Safer Trident that's powered by Nvidia Jetson is the world's only semi autonomous vehicle (SAV) for professional divers. It's hybrid and multi-purpose, and relies on optical sensors and visual flow detection. Under low light conditions, even without Sonar or Doppler Velocity Log systems, it can acheive high-accuracy localization for underwater mapping. The machine has a unique visual navigation system which is completely stealth.
Claire Peng has over 6 years of professional experience in the media industry, covering TV, newspaper and online media. She was once a reporter and producer for Fairchild Television based in Toronto Canada, and worked as an English news reporter for the Global Times in Beijing. She writes mainly about self-driving, companies investment, and the enterprise lab.
NYU Releases the Largest LiDAR Dataset to Help Urban Development
July 13th, 2017 News of the Day: Uber merges with Yandex, U.S. to simplify driverless car rules
Hyperloop One Completes Its First Successful Test Run
July 12th, 2017 News of the Day: Faraday to move its factory, Vizio files lawsuit against LeEco
July 11th, 2017 News of the Day: Tesla to triple its service, Porsche posts half-year sales record
Faraday Future Halts Construction of $1 billion Factory in Nevada
July 10th, 2017 News of the Day: China to become Cadillac’s largest market, Oregon adds EV rebate
July 7th, 2017 News of the Day: VW partners with Kuka, Tesla to build battery factory in Australia
- Tesla Reveals Specs of its New AI-Powered Full Self-Driving Computer
- GM Sells Lordstown Plant to EV Startup Lordstown Motors
- China’s BIAC Group to Launch a New ‘Intelligent Car’ Brand
- University of Michigan Researchers Examining Motion Sickness in Autonomous Cars
- Electric Vehicle Maker Workhorse to Unveil its Next-Gen Electric Delivery Van
- Waymo CEO Says it Might Deploy its Self-Driving Technology in the Trucking Industry
- Strong SUV Sales Nullifying Environmental Benefits From Electric Vehicles
- General Motors, Volkswagen Want to Go Straight to Fully Electric Cars
- GM Takes Heat for Backing the Trump’s Administration Plan to Revoke California’s Ability to Set its Own Emission Standards
- New Consortium Aims to Develop a Standardized Computing Platform for Autonomous Vehicles