MAP2: Model- and Acceleration-Based Pursuit with MPC and Gaussian Process Residual Learning for Autonomous Racing figure
AlphaXiv 中文概览(可滚动查看)