The final goal of this project is to develop a model of human 2D-motion perception. The percept of velocity is not only influenced by speed and direction of the moving object by a large number of factors like contrast, nature of the stimulus(grating vs. tiny dots vs. large dot etc), size, presence of noise etc. The approach has been to use the current psychophysical and neurophysiological data and constantly develop psychophysical experiments to further channel the hypothesis
Consider three clocks that keep the time properly. The first one is an ordinary analog clock, the second has a smoothly moving seconds hand, and the third one has a seconds hand that rotates by a large angle at first and then sets back to the correct position. Though all the clocks keep the correct time, if viewed for short durations the smooth moving clock appears to be moving faster. This is the motivation for the present study. We are developing psychophysical experiments to further quantify the relationship between the perceived velocity and the spatiotemporal frequency of the object.
Many pairs of spatial and temporal frequencies in a motion display can result in the same stimulus speed for a moving object, but nonetheless these different pairs of spatial and temporal frequencies can produce different speed percepts. In the present experiment a small square or disk object (subtending approximately 0.5 to 2.9 degrees) moves intermittently;i.e., it is always present on the screen (unlike flicker) but moves by a small amount at fixed temporal intervals. The spatial movement of these objects varied from approximately 1.5 minutes to 2.2 degrees, at temporal intervals ranging from 10 milliseconds to a second. (The subjects’ heads were not fixed;all calculations assume a nominal viewing distance of 60 cm.) Using the method of paired comparisons, the observer is asked to choose which of two objects is moving faster over its linear trajectory. Each object has a different combination of spatial and temporal frequencies that result in the same overall speed as the other object. Results for four observers indicate that the subjective perception of the speed of an object depends on the spatiotemporal frequency of the moving pattern. This estimated speed varies as an inverted U-shaped function, peaking at a specific spatial and temporal frequency combination and decreasing as one moves away from it. While individual differences are found,the location of this peak is independent of the size of the object that isbeing tracked and its shape. The presence of sharp rather than blurred object boundaries and the shape of the object did not have a noticeable effect on the perceived speed. In conclusion, the perceived speed of an object varies systematically across combinations of spatiotemporal frequency that produce equivalent stimulus speeds.