Figure 3.

Changes in flow parsing during self-motion. (a) Forwards translation over a simple ground plane produces a pattern of expanding optic flow (black lines). Observed visual motion of an object within the environment depends on a combination of simulated self-motion and object motion. (b) Shown is a zoomed-in view of the dashed grey box in (a). The flow parsing hypothesis posits that observers infer actual object motion (dashed red lines) by a vector subtraction of the estimated visual object motion due to self-motion (solid black lines) from observed object motion (solid red lines). The estimate of visual motion due to self-motion is larger in magnitude when an observer is moving (Right panel) compared to while stationary (Left panel), resulting in changes to inferred actual object motion (illustration of results from [129] based on schematic from [130]).