Abstract
Purpose: We have already reported that reversed-phi was perceived with Motion Defined Motion (MDM) stimulus (ARVO2001). This suggests that polarity-dependent detector function for MDM (Maruya & Sato, 2000; 2001: cf. Zanker, 1993). In this study, we further investigated the role of polarities by comparing perception characteristics for the stimuli with the same modulation waveform but containing only a single motion direction (uni-directional) and those with two opposite directions (bi-directional). The uni-directional motion was created by raising the modulation waveform for bi-directional motion by half of its amplitude. Methods: The stimulus was 10 (V) × 4 (H) deg horizontal motion-defined patterns consisted of dynamic random dots. Dot density was uniform over the whole field, and dots were moving horizontally (right or left). The dot speed was modulated along V-axis either by sinusoidal (Exp. 1) or missing fundamental (Exp. 2) function (spatial frequency: 0.2 c/d). In bi-directional condition, the speed was modulated either −100 to 100 or −200 to 200 min/sec (both to the right and left around 0). In uni-directional condition, the speed of the dots was modulated between 0 and 200 min/sec, i.e. by plus/minus 100 min/sec function raised by 100 min/sec DC level. A 5-frame apparent motion was presented with a 1/4-cycle shift between frames. SOA was fixed at 320 ms. Subject was asked to report the perceived direction by 2AFC. Results: We found the motion was perceived stronger under the uni-directional condition than under the bi-directional condition. This suggests, in the MDM detection, signals after local motion detection-level are processed separately depending on its polarity. If this is the case, in the bi-directional stimuli, the signals were processed by separate detectors depending on their polarities, so the waveform of local motion flow-field was rectified and the detection supposedly is less efficient than in coincident condition.