Local time CES Baidu Smart Car (Baidu Intelligent), the world's leading provider of intelligent automotive software solutions It is reported that Road It is understood that decision-making and planning in autonomous driving has always been one of the most difficult problems for manufacturers. The current advanced autonomous driving is mainly composed of positioning, sensing, decision-making, planning and other technologies to form a complete autonomous driving solution, and End Road However, because advanced autonomous driving involves big data, deep learning and other technologies, it requires a lot of manpower, material resources, and financial resources. Most developers who speak autonomous driving technology are difficult to contact with deep learning algorithm programmers. Such resource barriers are undoubtedly not conducive to the development of autonomous driving technology. Therefore, Baidu Smart Car launched Road Zhang Hui, general manager of HMI & Ecology of Baidu Smart Car Division, said Road According to Zhang Hui, for the opening of big data, Baidu Smart Car will open 10,000 kilometers of autopilot training data in the first phase, and will gradually open all the big data resources owned by the company for future use. In addition, Baidu Smart Car will also rely on Road According to the CES live news, many car manufacturers have said that they will accelerate the mass production of self-driving cars. In this regard, the industry believes that the cooperation between AI companies with intelligent technology and car manufacturers will further accelerate, and the technical demand in related fields will reach a climax. At present, only Baidu Smart Car has cooperated with more than 60 auto manufacturers and many smart chip manufacturers to achieve intelligent driving technology adaptation of more than 150 models. Zhang Yaqin, president of Baidu, said at CES that smart cars are still in their infancy at this stage, so they need a good ecological environment. And Baidu Smart Car released Road this time Changzhou Jinxin ARTS & CRAFTS Co.,Ltd , https://www.jxpartyconfetti.com
2017 was held in Las Vegas, USA. After the Master ignited the artificial intelligence, it became a hot topic covering CES. In this context, the large number of exhibitors of the automakers is the automatic in the context of artificial intelligence. Driving a car has become the focus.
Vehicle) After joining hands with Beijing Auto as a strategic partner, it immediately announced the launch of the advanced autopilot artificial intelligence model Road.
Hackers, said that Baidu's autopilot training data based on this model will be opened to the public, and the Department of Automotive Engineering of Tsinghua University will conduct a series of actions such as research and development of deep auto-driving technology for Chinese roads.
The Hackers platform is an advanced autopilot artificial intelligence model that uses the artificial intelligence technology of Baidu deep learning and Baidu neural network to realize the optimal algorithm under real road conditions. Also, Road
Hackers is the world's first advanced autopilot model that can implement End to end mode on real roads.
To
End mode is used as the input of the original big data of the camera, radar and other sensors as input, using some mathematical model algorithm to calculate the optimal driving decision, and directly output to the vehicle trajectory, speed, direction, etc., the driving decision method needed in the actual road condition It also faces the need to continuously optimize the model algorithm as it continues to fit the real road conditions.
Hackers is based on a large amount of real-time driving data exported by Baidu map and other resources. On the one hand, it optimizes reasonable driving behavior and habits. On the other hand, it combines Baidu's leading deep learning and neural network to establish a mathematical algorithm model that is continuously optimized. According to the actual road conditions, the real-time output is optimally driven to the vehicle. When the vehicle is driven automatically, it can achieve the best choice in terms of comfort, safety, and energy consumption.
At the same time, Hackers also synchronizes the opening of autonomous driving training data to developers who have the ability and no opportunity to use it, explore research and even test their own algorithms and models, in the hope of forming a collective force to optimize the depth of autonomous driving. Learn algorithms and models.
Hackers will gradually expand the resources of opening up to the outside world, and continue to reduce the research threshold of auto-driving related technologies and algorithms. Future developers can validate and simulate innovative smart driving technologies by focusing on the high-level structure of the model without worrying about any trivial underlying issues such as data volume and source.
Hackers, together with the well-known university of higher education in Tsinghua University, develops the autonomous driving technology in China's road environment, further strengthens the research and development and application of China's autonomous driving technology, and continuously builds Chinese intelligence with the core of technical resources and sharing. Car eco mode.
The Hackers platform and shared autonomous driving training data, as well as the original intention of the joint higher education institutions, are the necessary resource requirements for the development and development of smart car technology research, so that a large number of enterprises and talents can lower the threshold. Participating in the construction of the smart car ecosystem is of great significance for the development and popularization of the smart car industry.