On the eve of the Spring Festival, SenseTime Jueying and Dongfeng Motor jointly launched the industry’s first generative intelligent driving mass production solution, breaking through the limitations of traditional technical approaches and achieving a dual breakthrough in technological advancement and mass production feasibility. Currently, the technical paths in the intelligent driving industry show clear differentiation, with each facing insurmountable shortcomings. One approach is the two-stage end-to-end architecture, which separates perception and planning into two independent modules, resulting in natural information loss between modules, limiting decision-making efficiency and making it difficult to form an effective closed-loop problem-solving system. The other approach is imitation learning-based intelligent driving solutions, which can only replicate existing driving data and are easily at a loss when faced with unexpected long-tail scenarios. The industry’s first generative intelligent driving mass production solution developed by SenseTime Jueying and Dongfeng Motor avoids the flaws of these two paths, building a leading full-stack capability system and achieving a technological leap forward in the industry. (Shanghai Observer News)
View Original
This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.
Equipped vehicles with the "Super Brain," the first generative intelligent driving mass production plan announced
On the eve of the Spring Festival, SenseTime Jueying and Dongfeng Motor jointly launched the industry’s first generative intelligent driving mass production solution, breaking through the limitations of traditional technical approaches and achieving a dual breakthrough in technological advancement and mass production feasibility. Currently, the technical paths in the intelligent driving industry show clear differentiation, with each facing insurmountable shortcomings. One approach is the two-stage end-to-end architecture, which separates perception and planning into two independent modules, resulting in natural information loss between modules, limiting decision-making efficiency and making it difficult to form an effective closed-loop problem-solving system. The other approach is imitation learning-based intelligent driving solutions, which can only replicate existing driving data and are easily at a loss when faced with unexpected long-tail scenarios. The industry’s first generative intelligent driving mass production solution developed by SenseTime Jueying and Dongfeng Motor avoids the flaws of these two paths, building a leading full-stack capability system and achieving a technological leap forward in the industry. (Shanghai Observer News)