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

Table 3

Results obtained from various deep reinforcement learning models.

Deep RL method	Reward
PPO2	3855.0 ± 5.00
SAC	3756 ± 42.47
A2C	3455.0 ± 105.19
TD3	3054.0 ± 793.85