Learning from Demonstrations and Human Evaluative Feedbacks: Handling Sparsity and Imperfection Using Inverse Reinforcement Learning Approach

<div>(a) Grid world navigation domain. The white, blue, green, gray, and red cells depict the ground, puddle, grass, obstacle, and goal, respectively. The black circle represents the learner robot. (b) A snapshot of our highway car driving simulator.</div>

Journal of Robotics

fig2

Figure 2

Figure 2: Learning from Demonstrations and Human Evaluative Feedbacks: Handling Sparsity and Imperfection Using Inverse Reinforcement Learning Approach 

Figure 2 | Learning from Demonstrations and Human Evaluative Feedbacks: Handling Sparsity and Imperfection Using Inverse Reinforcement Learning Approach 