Baidu, US's Berkeley DeepDrive speed up self-driving research
China's internet giant Baidu's self-driving open platform Apollo announced on March 9 that it will join hands with Berkeley DeepDrive, an industrial alliance of the University of California, on deep learning and autonomous driving.
Baidu Apollo also launched the Apollo data and its technological brand Apollo Scape, officially opening its large-scale autonomous driving data set.
Apollo Scape data platform has 10 times more data than similar platforms, such as Cityscapes, including autonomous driving technology sensing, simulated scenarios, road network data and other high-resolution image data.
Baidu Apollo joins hands with Berkeley DeepDrive, an industrial alliance of the University of California, on deep learning and autonomous driving on March 9. [Photo provided to chinadaily.com.cn] |
Baidu said the platform covers more complex road conditions, so that up to 162 vehicles and 80 pedestrians can be included in a single image. The platform defines a total of 26 data instances with different semantic items, such as cars, bicycles, pedestrians, buildings, and street lamps, and will further cover more complex environments, weather, and traffic conditions.
Wang Haifeng, vice president of Baidu and president of the Baidu Research Institute, said, "The cooperation between Baidu and Berkeley DeepDrive will rely on Apollo's industrial resources and Berkeley's top academic team to accelerate the technological innovation, theoretical innovation, and application process of autonomous driving."
A line of Baidu Apollo's autonomous driving cars stop on a road. [Photo provided to chinadaily.com.cn] |
Berkeley DeepDrive is an industry alliance led by the University of California, Berkeley, for the application of cutting-edge technologies to computer vision and machine learning in the automotive field.
The alliance includes 20 autopilot companies around the world, such as Nvidia, Qualcomm, General Motors and Ford, and its research projects covers key technologies of autonomous driving, including sensoring, planning and decision-making, and deep learning.
A new Baidu Apollo self-driving car on stage in 2017 [Photo provided to chinadaily.com.cn] |