Sim-to-Real Reinforcement Learning for Autonomous Driving Using Pseudosegmentation Labeling and Dynamic Calibration

Figure 6

Results of linear regression between throttle and ∆n, which is the distance advanced toward the heading direction.