Motion Processing of Reverse Phi
MetadataShow full item record
Introduction: When the contrast of the successive images is reversed for a target in an apparent motion stimulus, the perceived direction is reversed, i.e., opposite of phi motion, which is an integration of the sequential position. This phenomenon is called reverse phi. It is believed to be processed through the interaction of ON and OFF pathways. We investigate the following aspects of this phenomenon. Aims: • Experiment 1 (Chapter 2): To measure the spatio-temporal characteristics and motion coherence thresholds of reverse phi and compared to those of phi motion. • Experiment 2 (Chapter 3): To measure the spatio-temporal characteristics and motion sensitivity at the central and peripheral presentations of reverse phi and phi motion • Experiment 3 (Chapter 4): To test the inhibition hypothesis of reverse phi motion using transparent motion stimuli. • Experiment 4 (Chapter 5): To investigate whether contrast reversals had an effect on the perceived speed. Methods: All experiments were conducted in 10 participants using random dot kinematograms (RDK). Reverse phi stimuli consisted of dots changing from one contrast polarity to another upon displacement. Phi stimulus maintained the same luminance polarity throughout the trial. • Experiment 1 (Chapter 2): The temporal intervals tested varied from 16.7 to 66.8ms in steps of 16.7ms. The spatial displacements tested ranged from 0.1 to 0.5 deg, and for 16.5ms and 33.4ms, and the displacements were extended to 1.35 deg. For motion coherence thresholds, the signal dots were varied from 0 to 100% in variable steps for phi and reverse phi motion. Subjects reported the direction of motion. • Experiment 2 (Chapter 3): RDK stimuli were presented at the fovea and the superior retina at 15deg eccentricity. The dot size was 0.13° for central stimulus, which was scaled up to 0.26° for peripheral presentations. The temporal intervals tested varied from 16.7 to 50.1ms in steps of 16.7ms. The spatial displacements tested ranged from 0.1 to 1.35 degrees. For motion coherence thresholds, the proportion of signal dots was varied from 0 to 100% with different step sizes for phi and reverse phi motion. Subjects reported the direction of motion. • Experiment 3 (Chapter 4): In the motion transparency experiment, two RDKs moved in the opposite directions at 100% coherence. The subjects had to report the direction of motion whether it was in the right diagonal or left diagonal direction. In the motion nulling experiment, phi and reverse phi motion moved in the opposite directions with a fixed number of reverse phi dots and varying number of phi dots. The subjects had to report the direction whether the dots moved in the left, right or in both directions. • Experiment 4 (Chapter 5): Two RDK stimuli were presented for 0.5 second each in a sequential order with an inter-stimulus interval of 200 ms. First interval contained the standard stimulus with one of the following speeds - 18 deg/s, 24 deg/s or 34 deg/s and the second interval contained the test stimulus, which was 50%, 70%, 100%, 120%, 145% and 200% of the standard speed. The subject’s task was to compare the speed of the two stimuli and indicate which of the two stimuli appeared to be faster. Four conditions were tested with phi and reverse phi motion being test and/or standard stimulus. Results: • Experiment 1 (Chapter 2): The optimal spatial offset in sequential images for reverse phi and phi motion was 0.3 to 0.5 deg. The optimal temporal offset was 16.7ms for reverse phi and 16.7 or 33.4ms for phi motion. The average coherence threshold for reverse phi (25.9±6.7%) was higher than that of phi motion (14.5±3.2%), but the difference was not significant when stimulus parameters were considered. • Experiment 2 (Chapter 3): Reverse phi was observed both in central and peripheral presentations. There was no difference in the percentage correct responses between central and peripheral presentations for phi and reverse phi, except at 0.1 and 0.4 deg spatial offsets of 33.4ms temporal interval where reverse phi was perceived better at the periphery at 0.1 deg and at the center at 0.4deg. There was no difference in the motion coherence threshold between central and peripheral presentations for either phi or for reverse phi motion. • Experiment 3 (Chapter 4): Subjects confirmed perceiving a reversed direction for a reverse phi stimulus using single RDK. In motion transparency experiment, subjects reported perceiving motion along the direction of stimulus displacement for both motion stimuli. In the motion nulling experiment, reverse phi motion was dominated by a much smaller phi motion signal. • Experiment 4 (Chapter 5): The speed discrimination thresholds for phi motion were 5.8, 7 and 8 deg/sec for the standard stimuli of 18, 24 and 34 deg/sec, respectively. It was not possible to obtain a speed discrimination threshold for reverse phi motion because slower test speeds were perceived as faster than the standard speeds and vice versa except for the 18deg/sec standard speed, where faster test speeds were perceived as slower. When reverse phi and phi motion were compared, reverse phi was perceived as faster in 93.3%±4% of the trials when the phi motion was of the same speed as the reverse phi, despite changing the order of the presentation. Conclusions: • Experiment 1 (Chapter 2): The spatio-temporal characteristics of phi and reverse phi motion largely overlap. This indicates that a common mechanism, short-range system, processes the two types of motion. However, processing higher level tasks that involves segregation of signal from noise shows that reverse phi is less salient. • Experiment 2 (Chapter 3): Although there are anatomical and physiological differences between the center and periphery, the motion signals of reverse phi are processed equally well at the fovea and the retinal eccentricity tested. • Experiment 3 (Chapter 4): In reverse phi motion, transparency motion was perceived rather than an orthogonal motion. This suggests two possible conclusions: 1) there is no inhibition caused by a reverse phi motion on neurons tuned to the direction of physical displacement suggesting that reverse phi follows evidence-only hypothesis at the low-level motion detectors, 2) if any inhibition was present, it was insufficient to elicit an orthogonal motion. The results of nulling experiment suggest that reverse phi is a weaker stimulus in the presence of regular phi motion • Experiment 4 (Chapter 5): Slower speeds of reverse phi motion was perceived to be faster than the standard speeds due to the jerkiness inherent in the stimulus at slow speeds. The perceived speed of reverse phi was overestimated relative to phi motion when both were moving at the same speed. The overall results suggest that the spatio-temporal characteristics of reverse phi are similar to that of phi motion, however, reverse phi is a weaker stimulus resulting in a lower sensitivity. Motion transparency is possible with reverse phi and with phi motion. Reverse phi is perceived as being faster than phi motion especially at smaller displacements.
Cite this version of the work
Mohana Kuppuswamy Parthasarathy (2019). Motion Processing of Reverse Phi. UWSpace. http://hdl.handle.net/10012/15256