In the era of block worlds, the strategy to control was to first observe and then move. This was in general was called visual feedback. The term visual servoing appears to have been introduced by Hill & Park [Real time control of a robot with a mobile camera] in 1979, to distinguish from the earlier approach of first taking picture and then acting.

Object Centric Representation for Control

Learning Visual Predictive Models of Physics for Playing Billiards

Deep Object-Centric Representations for Generalizable Robot Learning