Abstract
A motion system transforms a function of x,y,t (e.g., the luminance input to one eye) into a motion vector flowfield. Lu and Sperling (1995, 2001) proposed that human visual motion perception involves three independent motion systems. The first-order system responds to moving luminance patterns—e.g., typical object motion. The second-order system detects moving modulations of feature activity. For example, it can detect the movement of stimuli in which the expected luminance is the same everywhere but an area of higher contrast moves. The third-order system computes the motion of figure-ground. Its input is a neural representation of visual space—a salience map—in which the locations of important visual features (figure) are marked (1) and “ground” is unmarked (0). Most stimuli excite all three systems.
The third-order computation is directly influenced by attention because attention influences salience. Attention to a color can be shown to increase its salience by about 30%. By embedding an attended color in an amplifying “sandwich” motion sequence, the apparent direction of motion becomes an extremely sensitive measure of attention. Several hours of searching for a particular color in a detection task influences subsequent tests of apparent motion immediately and continuing for about a month.
For an isoluminant red-green grating on a gray background, the salience of the red and green stripes is determined by the color difference from the background, i.e., by hue saturation. In a physically moving isoluminant red-green grating, equating the salience of the red and green stripes produces an apparently stationary grating, so-called “motion standstill”.
Motion standstill demonstrates that the object-shape-texture-color systems extract a single representative image from a sequence of rapidly changing retinal images. Even the perception a smoothly moving object is the combination of a representative image from object-shape-texture-color systems with the combined vector flowfield produced by the motion systems.
Lu, Z-L. & Sperling, G. (1995). The functional architecture of human visual motion perception. Vision Research, 35, 2697–2722.
Lu, Z-L. & Sperling, G. (2001). Three systems theory of human visual motion perception: review and update. J. Optical Soc. America A, 18, 2331–2370.