Depicts our biologically inspired proposal of visual motion processing that models the functions of primary visual area (V1), middle temporal area (MT), and medial superior temporal area (MST) as well as their interaction, indicated by an arrow.

<p>We consider cells from the dorsal part of area MST, referred to by MSTd, which respond to motion patterns. The three gray boxes show details of these functions. The box to the lower-left shows V1 detecting initial motions from Gabor filtering results at multiple spatial scales. The box to the upper-left shows the motion integration mechanisms of MT that encodes motion in neurons selective to a combination of motion speed and direction (velocity). This velocity space representation is displayed here in a Cartesian coordinate frame for convenience. A center-surround interaction is applied in these codomains of speed and direction. The box to the upper-right shows the globally defined motion patterns that extend over a large region of the visual field and that are detected in model area MSTd receiving input from MT.