Figure 11
From: Manipulation of free-floating objects using Faraday flows and deep reinforcement learning
A typical view returned from the camera, indicating the obscured zone, the shaded area in which starting positions are randomized, and the operating semicircle in which target positions are set.
