Eye against eye: How perceived and unperceived visual information interacts

Abstract

When the retinal images of the left and the right eye are incompatible, perception starts alternating. That is, the visual content that one is aware of alternates between the two images presented. This phenomenon is referred to as binocular rivalry. When one is aware of one of the two images, this image is referred to as dominant. The other image is then referred to as suppressed. In the current thesis, two unresolved issues that are central to understanding binocular rivalry are addressed. First, we focus on the interactions between aware and unaware visual information, a topic which is surrounded by many controversies. For example, there is evidence both in favor of and against binocular suppression as being non-selective. One of the key results of the present thesis is that suppression during binocular rivalry is feature-selective. Specifically, results indicate that suppression is not uniform across orientations: suppression was much stronger for orientations within a narrow range around the orientation of the suppressed rival grating, with a tuning similar reminiscent of orientation bandwidths found in V1 neurons. The same principle holds for spatial frequenty content. We go on to show that binocular rivalry competition not only affects suppressed selectively, but also the visual information of dominant perception. In other words, dominant images are affected by suppressed images in a very specific manner that can be related to the suppressed image’s feature content. Second, we discuss spatial interactions during rivalry. Although rivalry is a local phenomenon, spatial dependencies do exist. These dependencies reflect influences on binocular rivalry that may originate from different stage along the visual processing hierarchy. We go on to show that spatial dependencies in our experiments are primarily driven by the source (eye-of-origin) of the images. Image content (features) does affect spatial dependencies but this appears to be limited to the image’s most basic components, which are processed at early visual processing stages (i.e. primary visual cortex). These results suggest that higher-level image processing plays only a minor role in spatial dependencies during rivalry. The results presented in this thesis show strong interactions between aware and unaware visual information. These interactions are suggested to have an early, mainly monocular, locus. Likewise, the spatial dependencies seen for the outcome of rivalry competition (dominance) at multiple locations also appear to have a large early, monocular component. Moreover, our results suggest that higher-level modulation of dominance durations is rather weak. Taken together, the manipulations affecting perceptual suppression and dominance during binocular rivalry presented in this thesis will most likely affect processing at a monocular level of visual processing. In the context of the current dominant idea that rivalry is resolved at multiple stages of visual processing our studies provide evidence that the early stages are strongly involved, while higher-level areas contribute only little to resolving binocular rivalry.