GenAD: A groundbreaking end-to-end autonomous driving solution

The GenAD project, jointly developed by Berkeley, Huituo and the Institute of Automation of the Chinese Academy of Sciences, has led a new trend in the field of autonomous driving. The project proposes an innovative end-to-end autonomous driving paradigm, the core of which is to predict the evolution of the vehicle and its surroundings in future scenarios. GenAD uses a variational autoencoder to learn the distribution of future trajectories in the structural latent space, effectively realizing trajectory prior modeling. This project has been publicly released on GitHub, link: https://link.zhihu.com/?target=https%3A//github.com/wzzheng/GenAD