A computational model was developed to explain the effects of an interframe interval (IFI) in single-step apparent motion experiments. In these experiments a stimulus appears in one position, disappears, and then reappears in a shifted position after a short or long IFI. If the luminance during the IFI matches the mean luminance of the stimulus frames, long IFIs result in perceived motion opposite the short-IFI conditions. Brighter or darker IFIs, however, do not support the reversed motion effect. The model possesses the following defining characteristics: (1) a biphasic ("transient") channel whose signalled direction of motion reverses with changes of IFI duration; (2) a combined direction-opponent output which is the sum of directional responses developed in two channels -- biphasic ("transient") and monophasic ("sustained"); (3) a signal/noise weighting of the contributions of the two channels to the final directional output of the sstem. Predictions of the model about the effects of IFI intensity and viewing eccentricity were tested and confirmed in two new psychophysical experiments. The interpretations of past studies which included a role for second-order motion mechanisms in explaining IFI duration effects were reexamined. Further empirical tests of the model were outlined.
Keywords: