Dataset Information

Texture Interpolation for Probing Visual Perception.

ABSTRACT: Texture synthesis models are important tools for understanding visual processing. In particular, statistical approaches based on neurally relevant features have been instrumental in understanding aspects of visual perception and of neural coding. New deep learning-based approaches further improve the quality of synthetic textures. Yet, it is still unclear why deep texture synthesis performs so well, and applications of this new framework to probe visual perception are scarce. Here, we show that distributions of deep convolutional neural network (CNN) activations of a texture are well described by elliptical distributions and therefore, following optimal transport theory, constraining their mean and covariance is sufficient to generate new texture samples. Then, we propose the natural geodesics (i.e. the shortest path between two points) arising with the optimal transport metric to interpolate between arbitrary textures. Compared to other CNN-based approaches, our interpolation method appears to match more closely the geometry of texture perception, and our mathematical framework is better suited to study its statistical nature. We apply our method by measuring the perceptual scale associated to the interpolation parameter in human observers, and the neural sensitivity of different areas of visual cortex in macaque monkeys.

SUBMITTER: Vacher J

PROVIDER: S-EPMC9681139 | biostudies-literature | 2020 Dec

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Texture Interpolation for Probing Visual Perception.

Vacher Jonathan J Davila Aida A Kohn Adam A Coen-Cagli Ruben R

Advances in neural information processing systems 20201201

Texture synthesis models are important tools for understanding visual processing. In particular, statistical approaches based on neurally relevant features have been instrumental in understanding aspects of visual perception and of neural coding. New deep learning-based approaches further improve the quality of synthetic textures. Yet, it is still unclear why deep texture synthesis performs so well, and applications of this new framework to probe visual perception are scarce. Here, we show that ...[more]

PMID: 36420050

Similar Datasets

Project description:Segregation of information flow along a dorsally directed pathway for processing object location and a ventrally directed pathway for processing object identity is well established in the visual and auditory systems, but is less clear in the somatosensory system. We hypothesized that segregation of location vs. identity information in touch would be evident if texture is the relevant property for stimulus identity, given the salience of texture for touch. Here, we used functional magnetic resonance imaging (fMRI) to investigate whether the pathways for haptic and visual processing of location and texture are segregated, and the extent of bisensory convergence. Haptic texture-selectivity was found in the parietal operculum and posterior visual cortex bilaterally, and in parts of left inferior frontal cortex. There was bisensory texture-selectivity at some of these sites in posterior visual and left inferior frontal cortex. Connectivity analyses demonstrated, in each modality, flow of information from unisensory non-selective areas to modality-specific texture-selective areas and further to bisensory texture-selective areas. Location-selectivity was mostly bisensory, occurring in dorsal areas, including the frontal eye fields and multiple regions around the intraparietal sulcus bilaterally. Many of these regions received input from unisensory areas in both modalities. Together with earlier studies, the activation and connectivity analyses of the present study establish that somatosensory processing flows into segregated pathways for location and object identity information. The location-selective somatosensory pathway converges with its visual counterpart in dorsal frontoparietal cortex, while the texture-selective somatosensory pathway runs through the parietal operculum before converging with its visual counterpart in visual and frontal cortex. Both segregation of sensory processing according to object property and multisensory convergence appear to be universal organizing principles.

Project description:Our world is filled with texture. For the human visual system, this is an important source of information for assessing environmental and material properties. Indeed-and presumably for this reason-the human visual system has regions dedicated to processing textures. Despite their abundance and apparent relevance, only recently the relationships between texture features and high-level judgments have captured the interest of mainstream science, despite long-standing indications for such relationships. In this study, we explore such relationships, as these might be used to predict perceived texture qualities. This is relevant, not only from a psychological/neuroscience perspective, but also for more applied fields such as design, architecture, and the visual arts. In two separate experiments, observers judged various qualities of visual textures such as beauty, roughness, naturalness, elegance, and complexity. Based on factor analysis, we find that in both experiments, ~75% of the variability in the judgments could be explained by a two-dimensional space, with axes that are closely aligned to the beauty and roughness judgments. That a two-dimensional judgment space suffices to capture most of the variability in the perceived texture qualities suggests that observers use a relatively limited set of internal scales on which to base various judgments, including aesthetic ones. Finally, for both of these judgments, we determined the relationship with a large number of texture features computed for each of the texture stimuli. We find that the presence of lower spatial frequencies, oblique orientations, higher intensity variation, higher saturation, and redness correlates with higher beauty ratings. Features that captured image intensity and uniformity correlated with roughness ratings. Therefore, a number of computational texture features are predictive of these judgments. This suggests that perceived texture qualities-including the aesthetic appreciation-are sufficiently universal to be predicted-with reasonable accuracy-based on the computed feature content of the textures.

Project description:The transition from animal to plant proteins is booming, and the development of meat analogues or alternatives quickly progressing. However, the acceptance of meat analogues by consumers is still limited, mainly due to disappointing organoleptic properties of these foods. The objective of this study was to investigate possible relationships among structure, textural characteristics, consumer acceptance, and sensory evaluation of commercially available meat analogues. The microstructure and texture of 13 chicken analogue pieces and 14 analogue burgers were evaluated with confocal laser scanning microscopy (CLSM) and texture profile analysis (TPA). The moisture of the samples was related to cooking losses and release of liquid upon compression after cooking. Meat products were included as references. A sensory panel (n = 71) evaluated both flavour and texture characteristics. For the chicken analogue pieces, samples with more added fibres had a harder and chewier texture but were less cohesive. No other relations between composition and structure/texture could be found. In the sensory evaluation, lower hardness and chewiness were only seen in products with more fat. A lower sensory hardness was found to be related to the presence of small air pockets. For analogue burgers, there was no clear relation between composition and structure/texture. However, instrumentally measured hardness, chewiness, and cohesiveness correlated well with the corresponding sensory attributes, even though they could not be clearly linked to a structural feature. Next to this, fat content showed a clear correlation to perceived fattiness. CLSM images of burgers with high perceived fattiness showed large areas of fat. Therefore, the release of large fat pools from the meat was most likely responsible for the perception of this attribute. However, perceived fattiness was not related to liking, which was the case also for chicken analogue pieces. For both pieces and burgers, even if some of the measured textural attributes could be linked to the sensory profile, the textural attributes in question could not explain the liking scores. Liking was related to other aspects, such as meaty flavour and juiciness, which were not directly linked to compositional or textural features. Juiciness was not directly related to the moisture loss of the products, indicating that this attribute is rather complex and probably involves a combination of characteristics. These results show that to increase the appreciation of meat analogues by consumers, improving simple texture attributes is not sufficient. Controlling sensory attributes with complex cross-modal perception is probably more important.

Dataset Information

Texture Interpolation for Probing Visual Perception.

Publications

Texture Interpolation for Probing Visual Perception.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets