Are you an EPFL student looking for a semester project?
Work with us on data science and visualisation projects, and deploy your project as an app on top of Graph Search.
Common factors are ubiquitous in cognition, audition, and somato-sensation. Surprisingly, there seems to be no common factor in vision, i.e., visual tasks correlate very weakly with each other. Here, we show that there are likely many very specific factors in vision. First, we have previously shown that the Ebbinghaus illusion correlates very little with other spatial illusions (except the Ponzo illusion). Here, we measured illusion strength in the classic Ebbinghaus illusion with disks and compared it with the illusion magnitude of 18 versions of the Ebbinghaus illusion, for example, having squares rather than disks or moving rather than static disks. In addition, we asked observers to compare the size of two single disks. All versions of the Ebbinghaus illusion strongly correlated with each other but not with the single disks comparison task, even though the tasks are the same. Second, we previously presented 10 illusions, each with 4 different luminance conditions, including 2 iso-luminant ones. For all 10 illusions, the luminance conditions correlated highly with each other. Here, we re-analyzed the data and found that there were almost no correlations between the 10 illusions, except for the vertical and horizontal bisection illusion, which correlated strongly with each other for all 4 luminance conditions. In both bisection illusions, there are likely different neurons, sensitive to vertical and horizontal orientations, involved. Hence, variations of an illusion correlate highly with each other, however, it seems that each illusion makes up its own factor.
Olaf Blanke, Bruno Herbelin, Florian Lance, Hsin-Ping Wu, Sophie Jacqueline Andrée Betka, Estelle Nakul
, ,