Visual Attention Research Articles

Rethinking Attentive Object Detection via Neural Attention Learning.

Visual attention advances object detection by attending neural networks to object representations. While existing methods incorporate empirical modules to empower network attention, we rethink attentive object detection from the network learning perspective in this work. We propose a NEural Attention Learning approach (NEAL) which consists of two parts. During the back-propagation of each training iteration, we first calculate the partial derivatives (a.k.a. the accumulated gradients) of the classification output with respect to the input features. We refine these partial derivatives to obtain attention response maps whose elements reflect the contributions to the final network predictions. Then, we formulate the attention response maps as extra objective functions, which are combined together with the original detection loss to train detectors in an end-to-end manner. In this way, we succeed in learning an attentive CNN model without introducing additional network structures. We apply NEAL to the two-stage object detection frameworks, which are usually composed of a CNN feature backbone, a region proposal network (RPN), and a classifier. We show that the proposed NEAL not only helps the RPN attend to objects but also enables the classifier to pay more attention to the premier positive samples. To this end, the localization (proposal generation) and classification mutually benefit from each other in our proposed method. Extensive experiments on large-scale benchmark datasets, including MS COCO 2017 and Pascal VOC 2012, demonstrate that the proposed NEAL algorithm advances the two-stage object detector over state-of-the-art approaches.

FSwin Transformer: Feature-Space Window Attention Vision Transformer for Image Classification

FSwin Transformer: Feature-Space Window Attention Vision Transformer for Image Classification

Research on Non-Destructive Testing of Curtain Wall Hangings Based on Efficient Channel Attention Vision Transformer Network

Research on Non-Destructive Testing of Curtain Wall Hangings Based on Efficient Channel Attention Vision Transformer Network

A Comparative Study on Visual Attention, Verbal Engagement and Parental Behaviors During Shared Book Reading Using Scripted and Unscripted Story Books for Children with Autism Spectrum Disorder

Purpose: Shared book reading is nowadays considered as an important early language stimulation technique in the field of speech language pathology which encourages parent-child bonding and creates a conducive environment for language development. The parent-child interaction is important for a child with Autism Spectrum Disorder. This interaction along with adequate and contingent parental responses fosters early language development in children. Language-related parent behaviors promote children’s oral language skills by not only exposing the child to book-related vocabulary (e.g., labeling pictures, explaining word meanings, and linking words to everyday events), but also through explicit teaching of meaning-related skills related to the overall structure of the story (i.e., story grammar). Print-related SBR behaviors draw the child’s attention to print-related features of the book, for example, letter names and sounds.Methods: To better understand the relationship between children’s visual attention, verbal engagement and parent shared book reading behaviors during shared book reading tasks, this study included 10 preschoolers on the Autism Spectrum and their parents sharing an unfamiliar storybook including which are scripted and unscripted. Videos of the shared book reading interactions were transcribed and coded for visual attention, verbal engagement and parent behaviors using observational coding schemes. The frequency of visual attention, verbal engagement and parental SBR behaviors were compared between the scripted and unscripted story books using statistical tests-Wilcoxon signed rank test and Fisher’s Exact test.Results: Result reveals that there was no statistically significant difference between frequency of visual attention during unscripted story books and scripted books. Parental behaviors parameters pertaining to shared book reading were analyzed and found a significant difference in usage of book language is greater for scripted story books compared to unscripted books.Conclusions: The key findings shows the type of book (scripted or unscripted) have a significant impact and its contribution to the intervention of children with Autism Spectrum Disorder. Parental SBR behaviors can in turn contribute to the development of childhood language and preliteracy skills.

The Effect of Ad Exposure on Consumer Visual Attentions Depending on Time Lapse in the Context of eSports

PURPOSE This study examined how consumers' visual attention to ads during eSports media consumption varies over time.METHODS An experimental study with a single factor, three-level within subject experimental design was conducted, utilizing an eye-tracker to measure visual attentions, including fixation count and duration. Seventy-eight students from a national university in city B participated in the experiment. A repeated measures ANOVA was conducted using the open-source statistical program R to test the research hypothesis.RESULTS Both the fixation count and duration were highest for the first ad and then gradually decreased for the second and third ad.CONCLUSIONS It is recommended that eSports sponsors should consider differentiating ad pricing based on the order of exposure, then expose the first ad presented more frequently and for extended periods, and consider different shapes, colors, and movements to prevent adaptation to the initial allocation of attention.

Cognition and behaviour of reception in museum spaces

Abstract In recent years, there has been a growing interest in translation studies within museum settings, enabling museums to enhance visitor engagement and create a more meaningful visiting experience. Within cognitive translation studies, the use of eye tracking technology has emerged as a widely adopted method for investigating translators’ allocation of visual attention and cognitive resources. However, the application of eye tracking in museum translation studies has received limited attention. This study provides a critical review of the utilisation of eye tracking to examine visitors’ perception and reception in museum settings from both behavioural and cognitive perspectives. The article begins by selecting and categorising the reviewed papers. It then describes the methods employed for collecting and analysing eye movement data, followed by summaries and critical comments on the findings from the existing literature. This article further evaluates the reliability, validity, and practicality of existing eye tracking research in museum translation, and finally proposes the potential directions for future cognitive translation studies in museums.

Biological visual attention convolutional neural network for aircraft skin defect detection

Inspired by the biological visual attention mechanism, we propose a visual attention convolutional neural network to solve the problem of identifying aircraft skin, but showing that the aircraft s kin image collected by the camera is rarely illuminated, and the contrast between defects and background is low and difficult to identify. Firstly, the U-Net with the encoder-decoder structure is used to initially segment the skin image that are provided by Airlines. Then, the residual block is introduced into the U-Net to enhance the propagation ability of the features and extract more defect detail feature information. Finally, the visual attention mechanism is used to increase the weight of the defect area to reduce the influence of uneven illumination on the model. The experimental results show that the proposed model has better defect segmentation effects in both visual effects and objective evaluation indexes. Our algorithm not only enhances the accuracy and effectiveness of skin defect identification but also provides management personnel with a more relevant tool, promoting the intelligence and information of air maintenance. This contributes to excellent management support for aviation companies and related industries.

Spatial memory-augmented visual navigation based on hierarchical deep reinforcement learning in unknown environments

Visual navigation in unknown environments poses significant challenges due to the presence of many obstacles and low-texture scenes. These factors may cause frequent collisions and tracking failure of feature-based visual Simultaneous Localization and Mapping (vSLAM). To avoid these issues, this paper proposes a spatial memory-augmented visual navigation system that combines a vSLAM module, a conventional global planner module, and a Hierarchical Reinforcement Learning (HRL)-based local planner module. Firstly, a real-time vSLAM named Salient-SLAM is proposed to augment the performance of visual navigation. Salient-SLAM creates a navigation mapping thread by combining a saliency prediction model to build a navigation map that categorizes environmental regions as occupied, explored, or noticeable. Spatial memory that contains spatial abstraction and saliency information of the environment can be further formed by encoding navigation maps, which helps the agent determine an optimal path towards its destination. An open-sourced saliency dataset is proposed to train the saliency prediction model by mimicking the visual attention mechanism. Secondly, a HRL method is proposed to automatically decompose local planning into a high-level policy selector and several low-level policies, where the latter produces actions to interact with the environment. We maximize entropy and minimize option correlation in learning low-level policies, aiming at acquiring diverse and independent behaviors. The simulation results show that the proposed HRL method outperforms competitive baselines by 6.29–10.85 % on Success Rate (SR) and 3.87–11.1 % on Success weighted by Path Length (SPL) metrics. By incorporating the spatial memory, SR, and SPL metrics can be augmented by an average of 9.85 % and 10.89 %, respectively.

The use of the eye-fixation-related potential to investigate visual perception in professional domains with high attentional demand: a literature review.

Visual attention is a cognitive skill related to visual perception and neural activity, and also moderated by expertise, in time-constrained professional domains (e.g., aviation, driving, sport, surgery). However, the contribution of both perceptual and neural processes on performance has been studied separately in the literature. We defend an integration of visual and neural signals to offer a more complete picture of the visual attention displayed by professionals of different skill levels when performing free-viewing tasks. Specifically, we propose to zoom the analysis in data related to the quiet eye and P300 component jointly, as a novel signal processing approach to evaluate professionals' visual attention. This review highlights the advantages of using portable eye trackers and electroencephalogram systems altogether, as a promising technique for a better understanding of early cognitive components related to attentional processes. Altogether, the eye-fixation-related potentials method may provide a better understanding of the cognitive mechanisms employed by the participants in natural settings, revealing what visual information is of interest for participants and distinguishing the neural bases of visual attention between targets and non-targets whenever they perceive a stimulus during free viewing experiments.

Nonlinguistic Cognitive Functions of Mandarin Speakers With Poststroke Aphasia.

The purpose of this study was to examine the cognitive functions of Mandarin speakers with poststroke aphasia and to investigate the relationship between nonlinguistic cognitive deficits and the severity of aphasia. Twenty-three adults with aphasia resulting from left-hemispheric stroke and 23 adults matched for age and educational level completed a series of six nonlinguistic cognitive tests measuring nonverbal intelligence, short-term memory, visual selective attention, visual alternating attention, auditory selective attention, and auditory alternating attention. A standardized aphasia assessment (Concise Chinese Aphasia Test [CCAT]) was also conducted to evaluate the severity of aphasia. Data analyses examined cognitive functions by comparing task performance of the two groups and examining the relationship between scores on the cognitive tasks and aphasia severity based on a hierarchical regression analysis. The aphasia group scored significantly lower than the control group on all nonlinguistic cognitive tasks with large effect sizes (d = 0.95 ~ 1.54). Significant associations between different nonlinguistic cognitive tasks and CCAT subtests were observed. Results from the hierarchical regression analysis showed that auditory alternating attention was the only factor that significantly predicted aphasia severity based on CCAT overall scores after age and education level were taken into account. The findings align with prior research observing deficits in nonlinguistic cognition in individuals with aphasia. Implications for clinical practice and future research are discussed.

COGNITIVE CHARACTERISTICS AND SOME FACTORS AFFECTING THE SPRINT AND MEMORY COURSE PERFORMANCE OF ORIENTEERS

It is known that physical, psychological, and cognitive characteristics are important in orienteering. However, no studies have investigated the effects of cognitive characteristics on performance in practice. This study aimed to determine the effect levels of the different variables of elite male orienteers (i.e., success rate, year of sports experience, short-term memory, visual memory, attention, and concentration) on their performance in two orienteering courses (sprint and memory courses). A total of 36 voluntary elite male athletes who had been orienteering for at least two years, over 18, with a mean age of 24.58±4.85 years and sports experience of 7.83±3.56 years participated in the study. The Cognitrone, VISGED visual memory, d2 attention, and serial digit learning tests were applied to all participants, respectively. All participants ran the memory and sprint orienteering courses at 72-hour intervals in a randomized manner. Afterward, the effects of the athletes’ cognitive performance and some characteristics on the course completion time were examined. According to the results of the linear regression analysis, the model’s independent variables that affected the winning time of the sprint course were the success rate, short-term memory score, and concentration (R= .505, R2= .255, F=3.193, p=.003, p<.01). Among these variables, the variable that predicted the winning time of the sprint course the most was the success rate (B=-244.25, Beta=-0.473, P=.015, p<.05). Year of sports experience, success rate, short-term memory score, concentration, and cog-correct response time were the independent variables affecting the winning time of the memory course (R=.620, R2=.385, F= 3.374, p=.000, p<.001). Year of sports experience was the independent variable with the highest effect on the winning time of the memory course (B=-6.718, Beta=-0.369, p=.029, p<.05). The factors affecting the sprint and memory courses' winning time were examined, and it was concluded that the orienteers' years of sports experience and successes affected the course winning time. These findings suggest that orienteering race performance is more related to the experience of transferring them to orienteering and creating tactical processes rather than the cognitive characteristics of orienteers.

Attention biases the process of risky decision-making: Evidence from eye-tracking.

Attention determines what kind of option information is processed during risky choices owing to the limitation of visual attention. This paper reviews research on the relationship between higher-complexity risky decision-making and attention as illustrated by eye-tracking to explain the process of risky decision-making by the effect of attention. We demonstrate this process from three stages: the pre-phase guidance of options on attention, the process of attention being biased, and the impact of attention on final risk preference. We conclude that exogenous information can capture attention directly to salient options, thereby altering evidence accumulation. In particular, for multi-attribute risky decision-making, attentional advantages increase the weight of specific attributes, thus biasing risk preference in different directions. We highlight the significance of understanding how people use available information to weigh risks from an information-processing perspective via process data.

From traditional to digital: The impact of drones and virtual reality technologies on educational models in the post-epidemic era

Employment of various innovative tools for education in the post-epidemic era is an urgent issue of improving the quality and efficiency of education and the relevance of educational material. The study aims to investigate the impact of the use of drones and virtual reality in educational models of student learning in the post-pandemic period. The study was conducted using a quantitative method, a sociological survey, a pedagogical experiment, and mathematical processing of the results. The study involved fourth-year students (n=748) studying in the following specialties: geography and ecology (n=119), engineering, robotics (n=135), architecture and urban planning (n=120), agriculture, agronomy (n=118), information technology (IT) (n=139), journalism (n=117), and teachers (n=258) with more than 5 years of experience. For the educational process of students of each specialty, the scope of application of drones and virtual reality has been developed. Classes with the use of drones and virtual reality were held two times a week for one academic semester. Afterward, students were interviewed using an adapted questionnaire to study their attitudes toward virtual reality technologies and drones for learning. Students of all specialties positively evaluated various aspects of the use of drones and virtual reality in their educational process. The use of drones helps to improve student engagement and visual attention and has a positive impact on learning efficiency and student satisfaction with the learning process. Teachers approve of the use of drones and virtual reality in teaching but point out that further development of the methodology for training stakeholders in the educational process, infrastructure, improvement of curricula, and consideration of various aspects of technology use is needed to effectively implement drones and virtual reality in the learning process. The practical significance of this study is to present aspects of the use of drones in teaching students of certain specialties and to study the impact of the effectiveness of the use of these technologies on the opinions of students and teachers.

Postural stability assessment in expert versus amateur basketball players during optic flow stimulation

We evaluated the role of visual stimulation on postural muscles and the changes in the center of pressure (CoP) during standing posture in expert and amateur basketball players. Participants were instructed to look at a fixation point presented on a screen during foveal, peripheral, and full field optic flow stimuli. Postural mechanisms and motor strategies were assessed by simultaneous recordings of stabilometric, oculomotor, and electromyographic data during visual stimulation. We found significant differences between experts and amateurs in the orientation of visual attention. Experts oriented attention to the right of their visual field, while amateurs to the bottom-right. The displacement in the CoP mediolateral direction showed that experts had a greater postural sway of the right leg, while amateurs on the left leg. The entropy-based data analysis of the CoP mediolateral direction exhibited a greater value in amateurs than in experts. The root-mean-square and the coactivation index analysis showed that experts activated mainly the right leg while amateurs the left leg. In conclusion, playing sports for years seems to have induced some strong differences in the standing posture between the right and left sides. Even during non-ecological visual stimulation, athletes maintain postural adaptations to counteract the body oscillation.

Different neural mechanisms for nonsalient trained stimuli and physically salient stimuli in visual processing.

Previous studies have shown that nonsalient trained stimuli could capture attention and would be actively suppressed when served as distractors. However, it was unclear whether nonsalient trained stimuli and physically salient stimuli operate through the same attentional neural mechanism. In the current study, we investigated this question by recording event-related potentials (ERPs) of searching for the two stimuli separately after matching the difficulty. The present results provided additional evidence for the function of the suppression in that it may terminate a shift of attention. For the N1 component, the nonsalient trained stimuli had a shorter latency and larger amplitude than the physically salient stimuli whether presented as targets or distractors. It indicated that the nonsalient trained stimuli had an earlier sensory processing and greater visual attention orienting. The N2 posterior-contralateral (N2pc) amplitude of the physically salient target was larger than the nonsalient trained target. This suggested that physically salient stimuli had a stronger ability to capture attention. However, when they presented as distractors, only the nonsalient trained stimuli could elicit the PD component. Therefore, active suppression of the physically salient stimuli was more difficult than the nonsalient trained stimulus with the same difficulty. For the P3 component, the amplitude of the physically salient stimuli was larger than that of the nonsalient trained stimuli, both as targets and distractors, which indicated that the top-down controlled process of outcome evaluation for the salient triangle was stronger. Overall, these results suggested that they were processed via different neural mechanisms in the early sensory processing, attentional selection, active suppression, and the outcome-evaluation process.

The Role of Attractiveness in Gendered Sexual Response Patterns

ABSTRACT Previously documented sexual response patterns of gender-specificity among gynephilic men and gender-nonspecificity among gynephilic women could be explained by women responding more strongly to non-gendered aspects of sexual stimuli. Cues of attractiveness are known determinants of sexual decision-making, yet have not been directly tested as determinants of sexual response. The current study investigated the role of attractiveness cues in explaining gender-based patterns of sexual response. Thirty-one gynephilic men and 60 androphilic women were presented slideshows of images depicting individual nude men and women that were pre-rated in a pilot study as either attractive or unattractive. The men and women were posed with legs spread and aroused genitals displayed prominently. Images were isolated against a white background and included minimal contextual information. Three sexual responses – genital arousal (via photoplethysmographs), self-reported arousal, and visual attention (via eye-tracking) – were recorded continuously. Across all three response modalities, men’s and women’s responses were stronger for the attractive versus unattractive images and for their preferred versus non-preferred gender. For men’s arousal and women’s self-reported arousal, the effect of attractiveness was stronger for their preferred versus non-preferred gender. Thus, both men and women demonstrated preference-specific patterns of sexual response. Gender cues had the strongest effect on men’s visual attention, whereas attractiveness cues had the strongest effect on women’s visual attention. Findings establish the importance of target attractiveness in arousal to sexual stimuli and add to mounting evidence that androphilic women’s sexual responses are sensitive to gender, but may be more sensitive to non-gendered features of sexual stimuli.

Visual attention towards building façades in the most energy vulnerable neighbourhoods in Madrid and the relaxing effect of visual coherence

ResumenEste estudio forma parte del proyecto mateMad que tiene como objetivo optimizar los materiales urbanos para hacer que los barrios desfavorecidos sean más habitables y sostenibles. El objetivo de este estudio preliminar es analizar la atención visual y el compromiso humano-fachada hacia las fachadas de edificios en barrios desfavorecidos en Madrid con el fin de crear barrios hermosos, sostenibles e inclusivos. La falta de simetría a lo largo del eje vertical y un tamaño geométrico no similar de las ventanas, mantenimiento deficiente, falta de transparencia de las ventanas y homogeneidad en color y textura resultaron en una mayor angustia subjetiva reportada, mientras que la simetría horizontal/vertical y la misma forma/tamaño de ventana resultaron en menos angustia. AbstractThis study is part of the mateMad project that aims to optimize urban materials to make deprived neighbourhoods more liveable and sustainable. The objective of this preliminary study is to analyse visual attention and human-façade engagement towards building façades in deprived neighbourhoods in Madrid to the end of creating beautiful, sustainable, and inclusive neighbourhoods. A lack of symmetry along the vertical axis and a-similar geometrical size of windows, poor maintenance, non-transparency of windows, and non-homogeneity in colour and texture resulted in more reported subjective distress, whereas horizontal/vertical symmetry and same window shape/size resulted in less.

Visual Attention and ODE-inspired Fusion Network for image dehazing

Image dehazing is an improtant image pre-processing step for many computer vision tasks with many proposed methods using convolutional neural networks. Ordinary Differential Equations (ODE) are a powerful mathematical tool in social and natural science, leading to better understanding, prediction, and use of information systems. Because prior knowledge has demonstrated its effectiveness in many practical domains, we design a Visual Attention Network (VAN) and an ODE-inspired Network (ODEN) based on prior knowledge. Then, we develop a multi-model fuzzy fusion strategy, which integrates results predicted by the prior-based Visual Attention Network (VAN) and the ODE-inspired Network (ODEN) into a single network to leverage their respective strengths in improving the dehazing performance. First, to utilize the haze-related prior for dehazing, the VAN removes haze with the help of the haze attention map. Then, to improve performance of stacked residual blocks inspired by the first-order Euler method in ODE-based methods, the ODEN is built by only 3 Runge–Kutta Modules (RKM), each of which not only is related to the stable fourth-order Runge–Kutta method implemented in the Runge–Kutta Block (RKB), but also combines the RKB with an attention mechanism. Finally, an attention based fusion mechanism is used to fuse results estimated by the ODEN and the VAN based on the multi-level features extracted by a pretrained ResNeXt. The experimental results demonstrate that the proposed method outperforms existing state-of-the-art methods in terms of visual effects and accuracy. The average PSNR and SSIM on three public datasets are 31.36 and 0.9766, respectively, which are better than the compared state-of-the-art methods.

Electrophysiological study of visual processing in children with cochlear implants

Electrophysiological studies of congenitally deaf children and adults have reported atypical visual evoked potentials (VEPs) which have been associated with both behavioral enhancements of visual attention as well as poorer performance and outcomes in tests of spoken language speech processing. This pattern has often been interpreted as a maladaptive consequence of early auditory deprivation, whereby a remapping of auditory cortex by the visual system ultimately reduces resources necessary for optimal rehabilitative outcomes of spoken language acquisition and use. Making use of a novel electrophysiological paradigm, we compare VEPs in children with severe to profound congenital deafness who received a cochlear implant(s) prior to 31 months (n = 28) and typically developing age matched controls (n = 28). We observe amplitude enhancements and in some cases latency differences in occipitally expressed P1 and N1 VEP components in CI-using children as well as an early frontal negativity, N1a. We relate these findings to developmental factors such as chronological age and spoken language understanding. We further evaluate whether VEPs are additionally modulated by auditory stimulation. Collectively, these data provide a means to examine the extent to which atypical VEPs are consistent with prior accounts of maladaptive cross-modal plasticity. Our results support a view that VEP changes reflect alterations to visual-sensory attention and saliency mechanisms rather than a re-mapping of auditory cortex. The present data suggests that early auditory deprivation may have temporally prolonged effects on visual system processing even after activation and use of cochlear implant.

Executive functions-based reading training engages the cingulo-opercular and dorsal attention networks.

The aim of this study was to determine the effect of a computerized executive functions (EFs)-based reading intervention on neural circuits supporting EFs and visual attention. Seed-to-voxel functional connectivity analysis was conducted focusing on large-scale attention system brain networks, during an fMRI reading fluency task. Participants were 8- to 12-year-old English-speaking children with dyslexia (n = 43) and typical readers (n = 36) trained on an EFs-based reading training (n = 40) versus math training (n = 39). Training duration was 8 weeks. After the EFs-based reading intervention, children with dyslexia improved their scores in reading rate and visual attention (compared to math intervention). Neurobiologically, children with dyslexia displayed an increase in functional connectivity strength after the intervention between the cingulo-opercular network and occipital and precentral regions. Noteworthy, the functional connectivity indices between these brain regions showed a positive correlation with speed of processing and visual attention scores in both pretest and posttest. The results suggest that reading improvement following an EFs-based reading intervention involves neuroplastic connectivity changes in brain areas related to EFs and primary visual processing in children with dyslexia. Our results highlight the need for training underlying cognitive abilities supporting reading, such as EFs and visual attention, in order to enhance reading abilities in dyslexia.