Perceptual completion supports transsaccadic object correspondence

Poster Presentation: Sunday, May 18, 2025, 8:30 am – 12:30 pm, Pavilion
Session: Eye Movements: Saccades, remapping

Fabian Parth1, Alexander C. Schütz1,2; 1Sensorimotor learning unit, University of Marburg, Germany, 2Center for Mind, Brain and Behavior, Marburg & Gießen, Germany

Saccadic eye movements bring objects from the peripheral visual field to the fovea for further inspection. To combine presaccadic with postsaccadic information, specific features—such as shape, texture, displacement, size, and contrast—are used to establish transsaccadic object correspondence (TOC). However, during saccades, external factors like partial occlusion can alter object information, raising the question of how the visual system copes with such changes and whether perceptual completion supports TOC. We conducted a postsaccadic target selection task to explore the dynamics of TOC and examine whether observers can establish TOC with partially occluded and amodally completed objects. Observers initiated a horizontal saccade towards a presaccadic peripheral object (circle or pacman), which was replaced perisaccadically by two vertically displaced candidates. These candidates were either fully visible and identical to the presaccadic object, partially occluding a triangle, or partially occluded by the triangle. Various combinations of candidates were tested to isolate the effects of shape, occlusion, configuration complexity, and other factors. At the end of each trial, observers had to indicate which postsaccadic candidate matched the presaccadic object. Behavioral responses were almost perfect and observers performed corrective saccades to the corresponding candidates with an accuracy of 80%. This indicates that TOC was possible even with dynamic changes in the available object information. Notably, corrective saccade latencies were longer when targeting candidates overlapping with the triangle, reflecting a cost for processing more complex configurations. Furthermore, latencies were shorter when perceptual completion facilitated the detection of the partially occluded target, compared to configurations where completion hindered selection. These findings suggest that perceptual completion supports TOC by resolving ambiguities in scenarios with partial occlusion. This highlights the visual system’s ability to integrate incomplete information and underscores the role of perceptual completion in maintaining object continuity across saccades.

Acknowledgements: FP and ACS were supported by Deutsche Forschungsgemeinschaft (DFG), Project number 290878970-GRK 2271/Project 5.