Purchase this article with an account.
Elif M. Sikoglu, Lucia M. Vaina; Effect of directional noise on heading perception. Journal of Vision 2005;5(8):839. doi: 10.1167/5.8.839.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Watamaniuk et al. (1989) demonstrated in a 2-D direction discrimination task that the difference between mean and perceived directions was modulated by a specified range of directionally constrained noise. Here we are interested to learn whether a similar effect could be observed in a heading perception task.
The experiment consisted of a random dot kinematograms (RDK) of expanding motion that produces the illusion of straight line heading shown at 44×44 deg2 aperture for 482 ms.. At the end of the motion, the RDK was replaced by a static random dot frame with the same statistical properties as the RDK and a vertical line. Observers were asked to determine whether heading direction was to the left or right of the vertical line. Adaptive staircase procedures were used to measure the accuracy of the heading perception (at 79%-correct level). Three experimental conditions differing in the directional noise perturbations were employed: 1) Random-walk: Each dot was perturbed independently from its direction in the previous frame; 2) Fixed-trajectory: Each dot kept the same perturbed direction throughout its lifetime; 3) General Perturbation: Location of heading direction was perturbed frame by frame.
In heading perception without noise, observers' accuracy was about 2°. Accuracy dropped to roughly 10°, when range of perturbation was 87°±4.6° for random-walk and 56°±8.5° for fixed-trajectory conditions. As the amount of perturbations increased, thresholds systematically increased in all three conditions (p < 0.05, slope of linear fits to perturbation range vs. threshold).
The results indicate that in the presence of either local (conditions 1 and 2) or global (condition 3) directional noise, heading direction can still be perceived. This implies that under noisy conditions both temporal and spatial integration mechanisms may help heading perception, suggesting that precise local direction perception is not required for the task.
This PDF is available to Subscribers Only