Interactions between Local and Global Signals in Global Motion Flow Processing

Abstract

Visual motion signals are processed through a hierarchical system consisting of multiple levels. The current thesis systematically examined global motion processing functions, from local motion integration to global motion interpretation, attempting to draw connections between the different motion processing stages. The initial experiment measured the global direction aftereffect (DAE) tuning using a novel stimulus that attempted to reduce the local motion adaptation. Compared with the control condition, the test adaptor displayed weaker, but significant global DAE, suggesting that DAE could occur at both local and global levels. Three further experiments examined the relationship between global motion adaptation and local motion orientation, based upon the assumption that a stronger MAE should be expected when adaptor and test share similar local motion orientations. The translational motion adaptation experiment reported that changing the local motion textural orientation did not result in reduced overall MAE. In direct contrast, the third and fourth experiments reported that, for different complex global motion adaptors, changing the local texture orientation did result in significantly reduced MAE. This contradiction suggests that there might be a top-down feedback pathway that affects local motion processing mechanisms differently, according to different types of global motion flow adaptors. The final experiment measured the sensitivity of global motion centre-of-flow (COF) detection. Although all participants showed better COF sensitivity for optic-flow related motion, different participants displayed different COF sensitivity preferences for different global and local motion cues; this likely reflects the fact that complex global motion flow interpretation is based upon a complex interplay of both local and global motion cues. The results of the studies described in this thesis suggest that the motion processing system consists of complex interactions between different stages in which there is top-down feedback from global motion processes, which may suppress local motion processing functions.

Date of Award	27 Sept 2022
Original language	English
Awarding Institution	University of Bristol
Supervisor	Christopher P Benton (Supervisor)