The microstructure of motion correspondences revealed by the Ternus-Pikler display

Michael Herzog, Laboratory of Psychophysics, Ecole Polytechnique Fédérale de Lausanne (EPFL), Switzerland

The Gestaltists were the first to investigate the rules of grouping by asking observers about subjective percepts. Here, we propose an indirect approach. We investigated the motion correspondence problem using the Ternus-Pikler display. Three lines were presented, followed by an ISI, and then moved by one line spacing. Apparent motion is perceived and a non-retinotopic reference is established between lines. Interestingly, a Vernier offset presented at the central line in the first frame is perceived at the central line of the second frame- according to the unambiguous, non-retinotopic line-to-line mapping. To understand the microstructure of motion correspondences, we tested Vernier offset attribution as a marker for grouping in displays with ambiguous correspondences, for example, by presenting first three and then four lines. We found clear-cut Vernier offset attribution between outer lines suggesting that these play a critical role in motion grouping, which is otherwise difficult to access using subjective reports.