Abstract
Texture may provide important clues for real world object and scene perception. To be reliable, these clues should ideally be invariant to common viewing variations such as changes in illumination and orientation. In a large image database of natural materials, we found textures with low-level contrast statistics that varied substantially under viewing variations, as well as textures that remained relatively constant. This led us to ask whether textures with constant contrast statistics give rise to more invariant representations compared to other textures. To test this, we selected natural texture images with either high (HV) or low (LV) variance in contrast statistics and presented these to human observers. In two distinct behavioral categorization paradigms, participants more often judged HV textures as “different” compared to LV textures, showing that textures with constant contrast statistics are perceived as being more invariant. In a separate electroencephalogram (EEG) experiment, evoked responses to single texture images (single-image ERPs) were collected. The results show that differences in contrast statistics correlated with both early and late differences in occipital ERP amplitude between individual images. Importantly, ERP differences between images of HV textures were mainly driven by illumination angle, which was not the case for LV images: there, differences were completely driven by texture membership. These converging neural and behavioral results imply that some natural textures are surprisingly invariant to illumination changes and that low-level contrast statistics are diagnostic of the extent of this invariance.
Highlights
Despite the complexity and variability of everyday visual input, the human brain rapidly translates light falling onto the retina into coherent percepts
0.1 0.2 0.3 0.4 0.5 0.6 same-texture standard deviation www.frontiersin.org images from the same HV texture category as different than vice versa. This finding suggests that LV texture categories are easier to categorize than HV texture categories and that images from the same HV texture category are perceived as less similar. This latter conclusion is supported by an additional analysis performed on the accuracy scores, in which we correlated the specific amount of variance in contrast statistics with the average number of same-texture errors
In a large database of natural textures, we selected images with low-level contrast statistics that were either constant or variable under changes in illumination angle and orientation. In both EEG and behavior, we showed that textures with little variation in low-level contrast statistics were perceived as more invariant and led to more invariant representations at the neural level
Summary
Despite the complexity and variability of everyday visual input, the human brain rapidly translates light falling onto the retina into coherent percepts. One of the relevant features to accomplish this feat is texture information (Bergen and Julesz, 1983; Malik and Perona, 1990; Elder and Velisavljevic, 2009). Texture—“the stuff in the image” (Adelson and Bergen, 1991)—is a property of an image region that can be used by early visual mechanisms for initial segmentation of the visual scene into regions (Landy and Graham, 2004), to separate figure from ground (Nothdurft, 1991) or to judge 3D shape from 2D input (Malik and Rosenholtz, 1997; Li and Zaidi, 2000). The relevance of texture for perception of natural images is demonstrated by the finding that a computational model based on texture statistics accurately predicted human natural scene categorization performance (Renninger and Malik, 2004). It can be hypothesized that textures that are more invariant will provide more reliable cues for object and scene perception
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
