Visual Properties Research Articles

Knowledge Guided Transformer Network for Compositional Zero-Shot Learning

Compositional Zero-shot Learning (CZSL) attempts to recognise images of new compositions of states and objects when images of only a subset of state-object compositions are available as training data. An example of CZSL is to recognise images of peeled apple by a model when it is trained using images of peeled orange , ripe apple and ripe orange . There are two major challenges in solving CZSL. First, the visual features of a state vary depending on the context of a state-object composition. For example state like ripe produces distinct visual properties in the compositions ripe orange and ripe banana . Hence, understanding the context dependency of state features is a necessary requirement to solve CZSL. Second, the extent of association between the features of a state and an object varies significantly in different images of same composition. For example, in different images of peeled oranges , the oranges may be peeled to different extents. As a consequence, the visual features of images of the class peeled orange may vary. Hence, there exists a significant amount of intra-class variability among the visual features of different images of a composition. Existing approaches merely look for the existence or absence of features of particular state or object in a composition. Our approach not only looks for the existence of a particular state features or object features but also the extent of association of state features and object features to better tackle the intra-class variability in visual features of compositional images. The proposed architecture is constructed using a novel Knowledge Guided Transformer . The transformer-based framework is utilised for processing larger context dependency between the state and object. Extensive experiments on C-GQA, MIT-States and UT-Zappos50k datasets demonstrate the superiority of the proposed approach in comparison with the state-of-the-art in both open-world and closed-world CZSL settings.

Diplomatic Communication of Silk Sign: Vehicle, Signification, and Narrative

ABSTRACT This study explores how silk as an aesthetic sign plays the active and efficient role of diplomatic narrative in the cultural and political communication between countries employing the historical and archaeological evidence in China. Aesthetic signs are of communicative values and cultural narrative significance. Silk is an aesthetic sign with iconic properties. The paper argues that beauty, exquisiteness, and value are the visual properties that silk as a sign vehicle presents in terms of texture, colours, patterns, and weaving techniques, recognised across all cultural contexts throughout history. The richness of favourable meanings that a sign conveys is embodied in the characteristics of its specific vehicle. The combination of silk as a sign vehicle and its designated object conveys the sign's value, thus performing the narrative function. Due to its iconic functions, silk is the most distinct and active aesthetic sign. Therefore, this paper argues that silk as a sign expresses diplomatic fraternity and harmony narrates aesthetics of unremitting pursuit of gentleness and auspiciousness, and constructs an international image of a country willing to coexist with others harmoniously and offer mutual assistance during difficulties.

Soft, crispy, crunchy, sustainable: The role of visual textures in shaping sustainable food preferences

This research evaluates the potential of using visual textural complexity as a subtle yet effective strategy to direct consumer preferences towards more sustainable food choices, addressing the urgent need to reduce greenhouse gas emissions in the food industry.Using a traditional Turkish rice pudding, sütlaç, as a case study, we prepared four variants with varying textural complexities (ranging from simple to complex) by reducing the portion size by half and gradually adding soft, crispy, crunchy, and airy textured food layers onto it. Two parallel sets of variants were formulated to minimize potential bias from differences in composition and visual properties of the layers. One hundred participants individually evaluated paired visual stimuli of these variants under two experimental conditions: non-informative, which included only visuals, and informative, which included visuals with accompanied by sample information.Paired binary analysis revealed consistent preference for variants with higher levels of visual textural complexity, especially those with three layers (soft, crispy and crunchy), regardless of experimental conditions and product formulations. Although sample information affected preference scores, variants with certain levels of textural complexity were still preferred over the traditional sample (p < 0.05). Preferences for sample variants over the control reduced the carbon emission over 31 %. The study links the impact of visual textural complexity on preferences to the Elaboration Likelihood Model (ELM) and demonstrates that this strategy can effectively direct consumer preferences towards more sustainable options. Adopting the insights from this study can assist food producers and marketers in contributing to the broader goal of reducing carbon emissions.

Computing Hive Plots: A Combinatorial Framework

Hive plots are a graph visualization style placing vertices on a set of radial axes emanating from a common center and drawing edges as smooth curves connecting their respective endpoints. In previous work on hive plots, assignment to an axis and vertex positions on each axis were determined based on selected vertex attributes and the order of axes was prespecified. Here, we present a new framework focusing on combinatorial aspects of these drawings to extend the original hive plot idea and optimize visual properties such as the total edge length and the number of edge crossings in the resulting hive plots. Our framework comprises three steps: (1) partition the vertices into multiple groups, each corresponding to an axis of the hive plot; (2) optimize the cyclic axis order to bring more strongly connected groups near each other; (3) optimize the vertex ordering on each axis to minimize edge crossings. Each of the three steps is related to a well-studied, but NP-complete computational problem. We combine and adapt suitable exact and heuristic algorithmic approaches, implement them as an instantiation of our framework, and show in a case study how it can be applied in a practical setting. Furthermore, we conduct computational experiments to gain further insights regarding the algorithmic choices of our framework. The code of the implementation and a prototype web application can be found on OSF.

Responsive integration performed by laser-induced Er3+ structural doping and graphene nanoplatelets

Laser-induced preparation method has the advantages of high efficiency, high yield, no ionic diffusion restriction and competitive side reaction, which can prepare diversified composite materials based on small-scale and distributed characteristics. Besides, integrating multiple performance and conducting inductive feedback are helpful for the reliability and calibration of laser-induced materials in the precision field. Herein, responsive integration is designed based on graphene nanoplatelets/ZnSe:Er3+ composite with optical, electrical, thermal and visual properties for real-time feedback. Er3+ ions doped ZnSe is obtained by fast laser-induced method, combined with two-dimensional graphene nanoplatelets and assembled into temperature-sensitive polyimide and cellulose paper, realizing the characteristics of photoelectric conversion and photothermal conversion. Besides, red fluorescence guiding perceptual behavior can also be produced under the radiation of NIR laser. Based on the proposed composite, bionic flower is constructed, and exhibits good motion features in blooming. Meanwhile, large curvature bending, high thermal sensitivity and strong fluorescence emission are also revealed. Such performance combination design provides a convenient approach for multi signals cross feedback in soft actuator, which strengthens the monitoring and guidance of motion perceptions, and can be potentially applied in diversified fields.

Intrinsic functional connectivity of right dorsolateral prefrontal cortex and hippocampus subregions relates to emotional and sensory-perceptual properties of intrusive trauma memories.

Trauma-related intrusive memories (TR-IMs), unwanted and vivid, are core symptoms of posttraumatic stress disorder (PTSD). Prior research links voluntary TR-IM suppression to inhibitory control of the right dorsolateral prefrontal cortex (dlPFC) over the hippocampus (HPC). However, the potential relevance of tonic resting-state inhibition has not been examined, nor has the functional differentiation of the anterior and posterior hippocampus (aHPC/HPC). This study examined relationships of TR-IM frequency and properties with resting-state negative coupling between the right dlPFC and right aHPC/pHPC in trauma-exposed individuals with PTSD symptoms. Participants (N=109; 88 female) completed two weeks of ecological momentary assessments capturing TR-IM frequency and properties (intrusiveness, emotional intensity, vividness, visual properties, and reliving). Using 3T resting-state magnetic resonance imaging, participant-specific 4-mm spheres were placed at the right dlPFC voxel most anticorrelated with the right aHPC and pHPC. Quasi-Poisson and linear mixed-effects models assessed relationships of TR-IM frequency and properties with right dlPFC-right aHPC and pHPC anticorrelation. TR-IM emotional intensity was positively associated with right dlPFC-aHPC connectivity, while vividness and visual properties were linked to right dlPFC-pHPC connectivity. No significant associations were found between TR-IM frequency, intrusiveness, or reliving, and anticorrelation with either HPC subregion. This study provides novel insights into the neural correlates of TR-IMs, highlighting the relevance of intrinsic negative coupling between the right dlPFC and aHPC/pHPC to their emotional impact and perceptual properties. Further research on inhibitory mechanisms in this circuit could improve understanding of component processes of intrusive reexperiencing, a severe and treatment-refractory PTSD symptom.

A Review on the Effect of Wood Surface Modification on Paint Film Adhesion Properties

Wood surface treatment aims to improve or reduce the surface activity of wood by physical treatment, chemical treatment, biological activation treatment or other methods to achieve the purpose of surface modification. After wood surface modification, the paint film adhesion performance, gluing performance, surface wettability, surface free energy and surface visual properties would be affected. This article aims to explore the effects of different modification methods on the adhesion of wood coating films. Modification of the wood surface significantly improves the adhesion properties of the paint film, thereby extending the service life of the coating. Research showed that physical external force modification improved the hydrophilicity and wettability of wood by changing its surface structure and texture, thus enhancing the adhesion of the coating. Additionally, high-temperature heat treatment modification reduced the risk of coating cracking and peeling by eliminating stress and moisture within the wood. Chemical impregnation modification utilized the different properties of organic and inorganic substances to improve the stability and durability of wood. Organic impregnation effectively filled the wood cell wall and increased its density, while inorganic impregnation enhanced the adhesion of the coating by forming stable chemical bonds. Composite modification methods combined the advantages of the above technologies and significantly improved the comprehensive properties of wood through multiple modification treatments, showing superior adhesion and durability. Comprehensive analysis indicated that selecting the appropriate modification method was key for different wood types and application environments.

A specialized inclusive road dataset with elevation profiles for realistic pedestrian navigation using open geospatial data and deep learning

Built environment characteristics can greatly influence pedestrians' route choices with factors beyond distance, such as accessibility, convenience, safety, and aesthetics, playing crucial roles. Although current navigation apps, such as Google Maps and Waze, have successfully provided driving directions, their navigation services are insufficient and sometimes unrealistic for addressing pedestrians' needs, largely due to the lack of dedicated pedestrian networks and the associated navigation algorithms. To address the research gaps, this paper proposes a novel approach that integrates freely available geospatial data and computer vision technology to create a specialized inclusive network dataset for outdoor pedestrian navigation. Moreover, a pedestrian navigation algorithm is developed to generate more practical “shortest” and “alternative” paths by incorporating various sidewalk attributes. We applied the method to create a pedestrian navigation network in Las Vegas. SpaceNet's open imagery dataset was used to extract Las Vegas's road networks. A virtual audit process assessed the visual and operational properties of the sidewalk networks using Google street-level images, evaluating factors including sidewalk presence, widths, surface types and conditions, missing curb ramps, greenery, protection from weather conditions, and lighting. Google Earth's open elevation data were used to analyze road elevation profiles as meaningful 3D indicators of sidewalk accessibility for wheelchair users. Further, additional geometric properties of the network, including road curviness, proximity to road intersections, and directional changes, were detected and analyzed. A navigation experiment conducted with individuals having varying mobility abilities, including regular pedestrians, older adults, and wheelchair users demonstrated the effectiveness of the newly developed network and algorithm in meeting the diverse needs of pedestrians.

Layered Modeling of Affective, Perception, and Visual Properties: Optimizing Structure With Genetic Algorithm

Layered Modeling of Affective, Perception, and Visual Properties: Optimizing Structure With Genetic Algorithm

A framework for automatically generating composite keywords for geo-tagged street images

Due to the ubiquity of sensor-equipped cameras such as smartphones, images are associated with spatial metadata, including camera’s geographical location and viewing orientation, which can be used for automatically generating better semantic keywords about geo-tagged urban street images in addition to visual keywords extracted from image analysis. This study introduces a novel framework for auto-tagging images that integrates both spatial and visual properties to generate comprehensive and accurate tags. The framework operates through four phases: extraction, abstraction, composition, and assessment. Our research highlights the benefits of combining visual and spatial analyses, demonstrated through a case study using geo-tagged urban street images from Orlando, Pittsburgh, and Manhattan. Experimental results show that the proposed framework significantly enhances the accuracy of keyword-based searches compared to conventional methods. In particular, based on our experiments, image search using the tags generated by our proposed framework, referred to as descriptive tags, achieved an average precision improvement factor of 0.9 compared to conventional tags. Additionally, our proposed ranking algorithm, which extends the term frequency-inverse document frequency (TF-IDF) algorithm, resulted in improvement factors of 0.86 for mean average precision (MAP) and 0.57 for mean reciprocal rank (MRR). Moreover, our framework’s flexibility and robustness make it suitable for diverse applications, from smart cities to online shopping. The paper also includes a detailed evaluation and user study, confirming the precision and reliability of the generated tags.

Image Restoration Technology of Tang Dynasty Tomb Murals Using Adversarial Edge Learning

The city of Xi’an, China houses a vast collection of valuable Tang dynasty (618 AD-907 AD) tomb murals that have experienced various degrees of damage over time due to weathering. This research proposes an adversarial edge learning-based mural restoration technique that provides a fast and accurate automatic repair of the murals. The method uses generative adversarial networks to learn how to repair the edge contour of a damaged mural and restore the missing content. The contour restoration network and mural restoration networks are trained and updated simultaneously to enhance the feature extraction and restoration coordination performance of the network. The experiments were conducted on a self-built Tang dynasty tomb mural painting restoration dataset and compared to several existing algorithms. The proposed algorithm showed a 1.3 increase in peak signal-to-noise ratio (PSNR) and a 1.43% increase in structural similarity (SSIM) value, providing a precise restoration of the structure and contour of the mural content. The algorithm successfully repaired damaged murals with fractures and small missing parts. The suggested method can be useful for both the technical repair and conservation of significant national visual cultural property and the digital restoration of historic murals.

Plant-based egg washes for use in baked goods: Machine learning and visual parameter analysis.

Pea protein is one potential environmentally sustainable way of recreating the functionality of eggs in coatings for baked goods. These coatings are commonly applied to enhance visual properties of baked goods that consumers desire, especially color and gloss. This project used a simplified pie crust formulation (flour, shortening, and water) as a model baked good system to investigate color and gloss properties of pea protein coatings comprising three pea protein glycerol ratios, prepared at 2pH values, and applied at three different weights, baked for either 15 or 30 min. Pictures were also taken and used to estimate L*, a*, and b* colorimeter values from images using random forest and artificial neural network models developed using SciKit-Learn. Image acquisition was also conducted at different lighting conditions and the results are corrected utilizing a principal component analysis (PCA) weighting approach. Results show that pea protein glycerol solutions are all glossier than egg washes, and the 80%/20% ratio gives the highest glossiness. Furthermore, there are not significant differences in any of the colorimeter values at any concentration when applying 1 g of protein. The artificial neural network model coupled with PCA weighting led to reasonable estimates of color at different lighting conditions. Together, these results offer an alternative for the use of pea protein/glycerol in lieu of egg washes.

Cross-channel adaptation reveals shared emotion representation from face and biological motion.

Emotions in interpersonal interactions can be communicated simultaneously via various social signals such as face and biological motion (BM). Here, we demonstrate that even though BM and face are very different in visual properties, emotions conveyed by these two types of social signals involve dedicated and common processing mechanisms (N = 168, college students, 2020-2024). By utilizing the visual adaptation paradigm, we found that prolonged exposure to the happy BM biased the emotion perception of the subsequently presented morphed BM toward sad, and vice versus. The observed aftereffect disappeared when the BM adaptors were shown inverted, indicating that it arose from emotional information processing rather than being a result of adaptation to constitutive low-level features. Besides, such an aftereffect was also found for facial expressions and similarly vanished when the face adaptors were inverted. Critically, preexposure to emotional faces also exerted an adaptation aftereffect on the emotion perception of BMs. Furthermore, this cross-channel effect could not only happen from faces to BMs but also from BMs to faces, suggesting that emotion perception from face and BM are potentially driven by common underlying neural substrates. Overall, these findings highlighted a close coupling of BM and face emotion perception and suggested the existence of a dedicated emotional representation that can be shared across these two different types of social signals. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Photo-modified and photo-degradable starch/Fe3O4/TiO2 nanocomposite: exploring the feasibility of reducing workforce by magnetic recycling.

Plastics are known for their durability and long decomposition time in the environment, which make plastic recycling an effective approach to mitigate plastic waste risks. However, the global plastic recycling rate is less than 10% mainly due to the labor-intensive and time-consuming nature of the manual recycling process, which poses high health risks and costs. Therefore, the development of a fast, effective, and operational process in current recycling plants is crucial to address the environmental concerns associated with plastics. In the current study, the feasibility of starch/Fe3O4/TiO2 bio-nanocomposite (SFT) as photo-modifiable and photo-degradable was investigated to reduce the workforce in recycling packaging material. The SFT was modified by different UV-C exposure times, which significantly altered its functional properties. The UV-C exposure increased the hydrophobicity of the SFT films and led to a homogenous distribution of Fe3O4/TiO2 nanoparticles (FT). It also increased tensile strength (TS) and decreased elongation at break (EB) of the films. It seems that producing shorter polymer chains, creating new linkages among the polymeric chains, and the homogenous distribution of FT in the matrix of biopolymer by UV-C are the main reasons for these changes. Moreover, the photo-degradation of SFT specimens increased significantly with longer UV-C exposure times. The utilization of magnetic properties in bio-based nanocomposites holds promising potential for streamlining labor-intensive processes in waste recycling plants. However, the inappropriate visual properties of SFT remain a significant obstacle that requires further attention to enable its commercial viability.

Neural substrates for saccadic modulation of visual representations in mouse superior colliculus.

How do sensory systems account for stimuli generated by natural behavior? We addressed this question by examining how an ethologically relevant class of saccades modulates visual representations in the mouse superior colliculus (SC), a key region for sensorimotor integration. We quantified saccadic modulation by recording SC responses to visual probes presented at stochastic saccade-probe latencies. Saccades significantly impacted population representations of the probes, with early enhancement that began prior to saccades and pronounced suppression for several hundred milliseconds following saccades, independent of units' visual response properties or directional tuning. To determine the cause of saccadic modulation, we presented fictive saccades that simulated the visual experience during saccades without motor output. Some units exhibited similar modulation by fictive and real saccades, suggesting a sensory-driven origin of saccadic modulation, while others had dissimilar modulation, indicating a motor contribution. These findings advance our understanding of the neural basis of natural visual coding.

A novel method for objectively classifying sequential emotion within dreams: a proof-of-concept pilot study.

Traditionally, emotions in dreams have been assessed using subjective ratings by human raters (e.g., external raters or dreamers themselves). These methods have extensive support and utility in dream science, yet they have certain innate limitations due to the subjective nature of the rating methodologies. Attempting to circumvent several of these limitations, we aimed to develop a novel method for objectively classifying and quantifying sequential (word-for-word) emotion within a dream report. We investigated whether sentiment analysis, a branch of natural language processing, could be used to generate continuous positive and negative valence ratings across a dream. In this pilot, proof-of-concept study, we used 14 dream reports collected upon awakening following overnight polysomnography. We also collected pre- and post-sleep affective data and personality metrics. Our objectives included demonstrating that (1) valence ratings derived from sentiment analysis (Valence Aware Dictionary for sEntiment Reasoning [VADER]) could be used to visualize (plot) positive and negative emotion fluctuations within a dream, (2) how the visual properties of emotion fluctuations within a dream (peaks and troughs, area under the curve) can be used to generate novel "emotion indicators" as proxies for emotion regulation throughout a dream, and (3) these emotion indicators correlate with sleep, affective, and personality variables known to be associated with dreaming and emotion regulation. We describe 6 novel, objective dream emotion indicators: Total number of Peaks, total number of Troughs, Positive, Negative, and Overall Emotion Intensity (composites from an "area under the curve" method using the trapezoid rule applied to the peaks and troughs), and the Emotion Gradient (a polynomial trendline fitted to the emotion fluctuations in the dream chart). The latter signifies the overall direction of sequential emotion changes within a dream. Results also showed that ⅚ emotion indicators correlated significantly with at least one existing sleep, affective, or personality variable known to be associated with dreaming and emotion regulation. We propose that the novel emotion indicators potentially serve as proxies for emotion regulation processes unfolding within a dream. These preliminary findings provide a methodological foundation for future studies to test and refine the method in larger and more diverse samples.

Optimization of stimulus properties for SSVEP-based BMI system with a heads-up display to control in-vehicle features.

Over the years, the driver-vehicle interface has been improved, but interacting with in-vehicle features can still increase distraction and affect road safety. This study aims to introduce brain-machine interface (BMI)- based solution to potentially enhance road safety. To achieve this goal, we evaluated visual stimuli properties (SPs) for a steady state visually evoked potentials (SSVEP)-based BMI system. We used a heads-up display (HUD) as the primary screen to present icons for controlling in-vehicle functions such as music, temperature, settings, and navigation. We investigated the effect of various SPs on SSVEP detection performance including the duty cycle and signal-to-noise ratio of visual stimuli, the size, color, and frequency of the icons, and array configuration and location. The experiments were conducted with 10 volunteers and the signals were analyzed using the canonical correlation analysis (CCA), filter bank CCA (FBCCA), and power spectral density analysis (PSDA). Our experimental results suggest that stimuli with a green color, a duty cycle of 50%, presented at a central location, with a size of 36 cm2 elicit a significantly stronger SSVEP response and enhanced SSVEP detection time. We also observed that lower SNR stimuli significantly affect SSVEP detection performance. There was no statistically significant difference observed in SSVEP response between the use of an LCD monitor and a HUD.

A fluorescence-validated MIP-ECL sensor based on UiO66 loaded carbon nitride for detection of trace Patulin in series fruit products

Trace-level mycotoxin patulin (PAT) exposure causes serious harm to human health. Herein, a fluorescence (FL)-validated molecularly imprinted polymer (MIP)-electrochemiluminescence (ECL) sensor based on zirconium metal organic frameworks loaded carbon nitride (UiO66@CN) for detection of trace PAT was proposed. UiO66@CN, with abundant-NH2 groups, excellent thermal and aqueous stability compared to CN, was prepared by a simple one-pot method. And it not only has excellent ECL performance, but also has reliable FL visualization properties. MIPs with 3D cavity containing specific PAT recognition sites were imprinted on UiO66@CN modified indium tin oxide by self-polymerization. The binding process of PAT molecules with MIP acted as a gate to control the FL and ECL signals. Notably, visualized FL and sensitive ECL were innovatively implemented in a sensing system. With the help of FL mode validation, PAT was accurately detected in the ECL mode with a detection limit of 50 fg mL−1. In addition, when proposed method and high-performance liquid chromatography (HPLC) were simultaneously applied to fruit products, the results obtained were in complete agreement. Thus, the assay provided a new approach to achieve low-cost, accurate and practical detection of trace toxins.

Effects of Moisture Content on the Radiative Properties and Energy-Saving Performance of Silica Aerogel Windows.

The thermal insulation performance of windows is crucial for energy-efficient buildings. Windows are typically the weakest part of the building envelope, regarding thermal insulation. Due to its excellent thermal insulation and high transparency, silica aerogel shows great promise as a window material. However, moisture can impact the effectiveness of the aerogel, leading to poor visibility and reduced thermal insulation. This study simulated a silica aerogel with varying moisture levels using the combination of diffusion-limited cluster aggregation, discrete dipole approximation, and Monte Carlo methods. The effects of the moisture content, thickness, porosity, and particle size on thermal conductivity, solar transmittance, and haze were analyzed. Visual properties of the aerogels were also considered. The energy consumption of a 30 m2 room under different climates was simulated using TRNSYS to assess the energy-saving potential of silica aerogel glass. The findings indicate that a higher moisture content leads to decreased solar transmittance and increased thermal conductivity of aerogels. Silica aerogel glass is more energy efficient than single-layer float glass, with the dry aerogel performing better in cold climates but worse in hot climates. This study provides insights for designing aerogel glass that optimizes solar transmittance and thermal insulation to enhance building comfort and energy efficiency.

In Vivo Optical Clearing of Mammalian Brain.

Established methods for imaging the living mammalian brain have, to date, taken optical properties of the tissue as fixed; we here demonstrate that it is possible to modify the optical properties of the brain itself to significantly enhance at-depth imaging while preserving native physiology. Using a small amount of any of several biocompatible materials to raise the refractive index of solutions superfusing the brain prior to imaging, we could increase several-fold the signals from the deepest cells normally visible and, under both one-photon and two-photon imaging, visualize cells previously too dim to see. The enhancement was observed for both anatomical and functional fluorescent reporters across a broad range of emission wavelengths. Importantly, visual tuning properties of cortical neurons in awake mice, and electrophysiological properties of neurons assessed ex vivo, were not altered by this procedure.