2017年第二届曲阜视觉科学会议知识库

PURPOSE: Orientation information (OI) is fundamental and crucial for our daily life. It can be processed as first-order luminance modulation, or it can be based upon higher level cues as contrast or texture, which are called second-order OI or higher. Our studies focus on the first-order OI processing and sought to answer the following two important questions: (1) dissociate first- and second-order orientation effects on the simple center-surround tilt illusion paradigm obtained with lateral flankers and (2) test whether the first-order spatio-orientation interactions follow the expected spatial segregation pattern of excitatory and inhibitory interactions (EI inter.) assumed in V1 models.
METHODS: We used as stimulus three Gabors patches whose centers stood in a straight line. By manipulating the bilateral Gabors’ orientations (flankers) and the global straight line orientation (envelope), and asking subjects to discriminate the orientation of the central target Gabor patch, we mapped the full two-dimensional spatio-orientation interaction structure onto the tilt repulsion effect of the central target Gabor. Seven adults participated in the experiment.
RESULTS: We observed a striking regular behaviour in the data. There was a repetitive pattern of flanker’s orientation effects onto perceived value across all global positions of the flankers, with the later 2nd order OI simply shifting vertically the reference point for 1st order OI effects. The assumed EI interactions of V1 models failed to predict our results.
CONCLUSIONS: Our data showed that the interaction of first-order elements, supposedly raised from the low-level visual cortex are created through spatial iso-orientation inhibitory interactions and that the second-order OI modulates these interactions in a global manner, without any first- and second-order interactions of OI.