Research Article

End-to-End Autonomous Exploration with Deep Reinforcement Learning and Intrinsic Motivation

Table 6

Experiment results of learning exploration with fine-tuning method (exist extrinsic reward).

EnvironmentMethodRewardMER (%)IQRE

Maze-1ICM + scratch584.59100.007.93
Ours + scratch586.32100.004.72
ICM + fine-tuning583.74100.009.13
Ours + fine-tuning586.56100.007.24
Maze-2ICM + scratch567.28100.008.07
Ours + scratch571.87100.005.15
ICM + fine-tuning514.6389.46N/A
Ours + fine-tuning569.44100.006.83
Maze-3ICM + scratch532.2791.64N/A
Ours + scratch579.65100.006.54
ICM + fine-tuning483.1682.95N/A
Ours + fine-tuning542.6892.63N/A