Figure 1 | Scientific Reports

Figure 1

From: Intrinsic interactive reinforcement learning – Using error-related potentials for real world human-robot interaction

Figure 1

Concept of the proposed approach. The robot tries to find an optimal action strategy through interaction with the human. The robot explores the possible action strategies and receives feedback (rewards) from the human. The goal of the robot is to maximize the total reward in the long run. In this way, the robot can learn and adapt its action strategy, while the human freely chooses the gestures and delivers feedback to the robot. In the end, the robot implicitly learns the meaning of human gestures.

Back to article page