Sight Research Articles

A Patch-Level Region-Aware Module with a Multi-Label Framework for Remote Sensing Image Captioning

Recent Transformer-based works can generate high-quality captions for remote sensing images (RSIs). However, these methods generally feed global or grid visual features to a Transformer-based captioning model for associating cross-modal information, which limits performance. In this work, we investigate unexplored ideas for a remote sensing image captioning task, using a novel patch-level region-aware module with a multi-label framework. Due to an overhead perspective and a significantly larger scale in RSIs, a patch-level region-aware module is designed to filter the redundant information in the RSI scene, which benefits the Transformer-based decoder by attaining improved image perception. Technically, the trainable multi-label classifier capitalizes on semantic features as supplementary to the region-aware features. Moreover, modeling the inner relations of inputs is essential for understanding the RSI. Thus, we introduce region-oriented attention, which associates region features and semantic labels, omits the irrelevant regions to highlight relevant regions, and learns related semantic information. Extensive qualitative and quantitative experimental results show the superiority of our approach on the RSICD, UCM-Captions, and Sydney-Captions. The code for our method will be publicly available.

Evaluation of a Visual Cognitive Style in Autism: A Cluster Analysis.

Cognitive style refers to how individuals perceive their environment and solve problems. Among various cognitive styles documented, verbal and visual styles, including object versus spatial cognition, have been widely studied. "Thinking in pictures" is often associated with autism and characterized by enhanced performance and autonomy of certain perceptual areas. This study aimed to characterize the cognitive style of autistic individuals, focusing on this visual "thinking in pictures" style. We assessed 43 autistic and 42 non-autistic adults using the Object Spatial Imagery Verbal Questionnaire to evaluate three dimensions of their cognitive style: visual object, visual spatial, and verbal. Using a cluster analysis, we identified cognitive style profiles. We then examined manifestations of cognitive style within these profiles, including mental imagery abilities (with the Mental Rotation Test), the vividness of mental images (with the Vividness of Visual Imagery Questionnaire), language abilities (with the Mill Hill vocabulary test), and synesthetic experiences. Our results revealed three distinct cognitive profiles: a visual profile characterized by significant mental imagery and vivid mental images, a visuo-spatial profile associated with strong abilities to manipulate mental images but lower vividness of mental images, and a verbal profile with lower spatial imagery and manipulation abilities. While all profiles were observed in autistic individuals, the visual cognitive profile was the most common. Additionally, we found synesthesia to be more prevalent in the autistic population compared to the non-autistic population. This study confirms that, while not exclusive, a visual cognitive style is prevalent among autistic individuals.

To analyse the impact of public relations strategies on the contribution of goals and initiatives of vision 2030

Background: This study aims to analyse the impact of public relations strategies on the promotion of Vision 2030 initiatives. Public relations, including media relations, crisis communication, and stakeholder engagement, are essential tools for shaping public perception and supporting national development plans. Method: A quantitative research approach is adopted for the study while the data collection technique employed is the structured questionnaires administered on a sample of 100 respondents. Descriptive analysis and correlation is done using statistical software known as SPSS in ascertaining the relationship between the proposed public relation strategies and the perception of Vision 2030 among the public. Findings: The study affirmed that management’s strategies of public relation have positively impacted public awareness and support to Vision 2030. Media relations and social media management enhance public communication. The stakeholders’ engagement is central to Vision 2030 goals promotion and success achievement. Recommendations: Media and social media strategies should be enhanced with the material of which the content is oriented in accordance with the designated demographic segment. Frequent and regular communication to the stakeholders would ensure that the news about the new Vision 2030 initiatives spreads widely hence getting support from many people. Moreover, the enhancement of crisis communication plans and protocols contributes to the increase in the public’s confidence and backing especially in turbulent times.

Integration Approaches in Forensic Portrait Examination

The article reviews some actual issues of the theory and practice of the production of forensic portrait examinations: the tight deadlines of the preliminary investigation, the increased time expenditures to produce forensic video examination, the need for an integrated approach to the training of forensic portrait experts. The author proposes to revise some approaches to the training of forensic experts in the field of forensic portrait examinations. Along with the possibility of a comprehensive forensic examination, taking into consideration current needs of practice, more attention needs to be paid to software application for processing photo and video images processing in order to improve visual perception of the information displayed and to study montage marks and their detectability. All this will contribute to the evolvement of a portrait examiner as a specialist of double competence.

WiTUnet: A U-shaped architecture integrating CNN and Transformer for improved feature alignment and local information fusion

Low-dose computed tomography (LDCT) has emerged as the preferred technology for diagnostic medical imaging due to the potential health risks associated with X-ray radiation and conventional computed tomography (CT) techniques. While LDCT utilizes a lower radiation dose compared to standard CT, it results in increased image noise, which can impair the accuracy of diagnoses. To mitigate this issue, advanced deep learning-based LDCT denoising algorithms have been developed. These primarily utilize Convolutional Neural Networks (CNNs) or Transformer Networks and often employ the Unet architecture, which enhances image detail by integrating feature maps from the encoder and decoder via skip connections. However, existing methods focus excessively on the optimization of the encoder and decoder structures while overlooking potential enhancements to the Unet architecture itself. This oversight can be problematic due to significant differences in feature map characteristics between the encoder and decoder, where simple fusion strategies may hinder effective image reconstruction. In this paper, we introduce WiTUnet, a novel LDCT image denoising method that utilizes nested, dense skip pathway in place of traditional skip connections to improve feature integration. Additionally, to address the high computational demands of conventional Transformers on large images, WiTUnet incorporates a windowed Transformer structure that processes images in smaller, non-overlapping segments, significantly reducing computational load. Moreover, our approach includes a Local Image Perception Enhancement (LiPe) module within both the encoder and decoder to replace the standard multi-layer perceptron (MLP) in Transformers, thereby improving the capture and representation of local image features. Through extensive experimental comparisons, WiTUnet has demonstrated superior performance over existing methods in critical metrics such as Peak Signal-to-Noise Ratio (PSNR), Structural Similarity (SSIM), and Root Mean Square Error (RMSE), significantly enhancing noise removal and image quality. The code is available on github https://github.com/woldier/WiTUNet.

Teachers’ Perception Towards Integration of Artificial Intelligence Tutoring-Based System in the School Curriculum: A Survey

Artificial Intelligence (AI) is the theory and development of computer systems capable of performing tasks normally requiring human intelligence, such as visual perception, speech recognition, decision-making, and translation between languages. This approach has been adopted by many organizations, including schools, to facilitate the teaching and learning process. This development has raised varying opinions on the future of employment within all sectors, including the educational sector. The purpose of this research was to evaluate the perception of teachers toward the integration of AI tutoring-based systems into the school curriculum. A descriptive analysis of the quantitative design was used, and 320 teachers were purposely sampled from public secondary schools in Southwest, Nigeria. The data were collected using a questionnaire and analyzed using SPSS. The findings of the study revealed that although teachers in public secondary schools in the southwest of Nigeria are aware of the positive impact of AI tutoring, 77% of the participants have a negative disposition towards the integration of an AI tutoring-based system into the school curriculum. The factors responsible for this high level of negative disposition include age, low level of education, fear of replacement and professional specialties of the teachers. Based on these findings, it was therefore recommended that teachers be sensitized on the importance of AI tutoring and that stakeholders should propagate the message of how it would promote the standard of education in the state. The study also recommended that educational policymakers integrate the use of artificial intelligence into the school curriculum by allocating the required resources to schools and organizing staff training on the use of AI without neglecting a befitting remuneration. Keywords: Artificial Intelligence, Teachers, Learning, Disposition, Perception

A preliminary exploration of quantitative models for evaluating binocular visual perception impairments among patients diagnosed with stroke.

The objective of this study is to formulate and implement graded biological models pertaining to binocular visual perception function with the use of computer algorithms. We aim to quantitatively assess the location, severity, and degree of impairment in binocular visual perception among patients who have suffered stroke, thereby providing valuable insights into the repercussions of cerebral tissue damage on the visual system. To overcome the shortcomings of previous instruments used to assess binocular function in terms of stereoscopic effects and the challenges posed by physiological and psychological interference during examinations, this study optimized its approach by integrating polarized stereovision technology with computer graphic modeling techniques. This study employed computer models to assess binocular visual perception function in stroke patients. Computer models refer to psychophysical testing methods used to measure binocular visual perception function, including various assessment tasks such as recognizing inverted letters and assessing stereopsis during high-speed movements. The cross-into-circle test was used as a means to quantify perceptual eye position. Subsequently, the collected data was analyzed to assess the magnitude of impairment in binocular visual perception. The results of the study revealed a spectrum of binocular visual perception impairment among patients diagnosed with stroke, demonstrating discernible variations in the recognition of inverted letters and stereopsis across different movement speeds. Importantly, perceptual eye position measurements offered valuable insights into ocular misalignment. The computational models effectively quantified both the spatial distribution and severity of these identified impairments. Damage to brain tissue resulting from a stroke can give rise to notable impairments in binocular visual perception function. Graded biological models, formulated through computer algorithms, provide a systematic framework for the comprehensive evaluation and quantification of these impairments. The comprehension of the nature and extent of visual impairments observed in patients with stroke establishes a basis for the development of personalized visual perception learning methodologies. Based on such tailored approaches, we aim to facilitate the recovery of impaired visual function, thereby contributing to the broader objective of neural system rehabilitation.

Self-Powered Infrared-Detectable BP/Ta2NiS5 Heterojunction and Its Application in Energy-Efficient Optoelectronic Synapses.

The development of energy-efficient and high-performance optoelectronic devices is crucial for the advancement of modern optoelectronic and microelectronic systems. Although the self-powered devices and optoelectronic synapses based on 2D heterojunction show great application prospects, the high energy consumption and infrared band detection of self-powered optoelectronic synapses are still an urgent problem to be solved. In this report, a BP/Ta2NiS5 heterojunction is constructed to achieve infrared detection by leveraging differences in Fermi energy levels. This heterojunction exhibits a high specific detectivity of 6.57×1010, 2.6×1010, and 1.12×1010 Jones and responsivity of 20, 10.6, and 5.9mAW-1 for 1064, 1550, and 2200nm infrared light at 0 bias voltage, respectively. In addition, under the 2200nm light, by applying an ultra-low bias voltage of 800µV, the heterojunction exhibits ultra-low power and energy consumption of 28.8pW and 0.64pJ, successfully simulates a variety of synaptic behaviors under infrared light, and demonstrates its image perception and image memory capabilities. These findings position the BP/Ta2NiS5 heterojunction as an ideal candidate for a multifunctional optoelectronic device crucial for advanced photodetection, neuromorphic computing, and artificial intelligence.

Traumatic Optic Neuropathy Treatment Trial 2 (TONTT-2): evaluating the efficacy of different doses of erythropoietin – a multicentre, randomised, double-blind clinical trial

AimThe aim is to compare the efficacy and safety of three different weight-adjusted intravenous erythropoietin (EPO) doses in patients with indirect traumatic optic neuropathy (TON).MethodsThis study is a multicentre, randomised,...

Body, belief, and postpartum recovery: perceptions of exercise, body image, and postpartum wellbeing amongst ultra-Orthodox women

ABSTRACT Exercise during the postpartum period positively influences mothers’ mental health, including body image, which is critical for a successful postpartum experience. Ultra-Orthodox women, subject to strict religious laws governing their bodies, strongly adhere to the commandment to procreate, resulting in a higher birth rate compared to the general population. These unique characteristics may find expression in the way religion shapes their postpartum body image and exercise experience. This study aims to explore these elements, addressing the scarcity of research on this population, to better understand unique cultural sensitivities during this critical period of mothers’ health. Semi-structured interviews with ultra-Orthodox postpartum women following 12 weeks of exercise, revealed: (1) The perception of the body among postpartum ultra-Orthodox women, including the body (1.1) perceived as a means, not a goal, (1.2) disrupted by pregnancy, and (1.3) healing through exercise. The second theme found (2) Exercise raises cultural conflicts, and participants (2.1) provided justification for exercise and (2.2) hiding participation. The discussion focuses on the interplay between religious beliefs, the postpartum female body, and internal conflicts arising in relation to the postpartum exercise experience. This qualitative study highlights the ways that these cultural nuances may advance or hinder postpartum wellbeing of ultra-Orthodox women.

Relationship between perception of body image on obesity and smoking status by age group in women: Findings of a seven-year Korean National Survey.

This descriptive study examined the relationship between body image perception and smoking status among women aged 19-64 years in Korea, using data from the Korea National Health and Nutrition Examination Survey (KNHANES). This study is a secondary analysis of data gathered from the KNHANES between 2014 and 2020, encompassing 12515 women aged 19-64 years. The final study group consisted of 742 current smokers (CS), 132 hidden smokers (HS), and 11641 non-smokers (NS). Hidden smokers were defined as participants who reported being non-smokers but had urine cotinine levels >50 ng/mL. The participants were divided into three age categories: 19-29, 30-49, and 50-64 years. A multiple logistic regression analysis was used to assess the relationship between body image perception and smoking status, by age group. Baseline statistics indicated that smokers (both CS and HS) generally had a lower socioeconomic status across all age groups. The highest rate of obesity perception was observed in the CS group, followed by the HS group, in both groups of women aged 19-29 and 30-49 years. However, only women aged 19-29 years in the CS group were more likely to perceive themselves as obese than those in the NS group (AOR=2.60; 95% CI: 1.49-4.52; p=0.001). Furthermore, factors such as current smoking status (AOR=2.32; 95% CI: 1.28-4.23; p=0.006), higher body mass index (AOR=2.95; 95% CI: 2.59-3.37; p<0.001), and perceived health status as poor (AOR=3.82; 95% CI: 2.11-6.92; p<0.001), significantly influenced the perception of obesity in this age group. This study identified a notable relationship between obesity perception and smoking among women aged 19-29 years only. These findings suggest that interventions aimed at weight reduction or modifying the perception of obesity, could potentially aid smoking cessation efforts in young women.

Robot intelligent perception and control

Abstract. In recent years, with the rapid development of robotics, intelligent perception and control has become the core research direction in the field of robotics. This technology plays a key role in many fields such as industrial automation, medical assistance, and autonomous driving, significantly improving production efficiency and service quality, and also promoting the intelligent processes of various industries. On the basis of an in-depth analysis of robot intelligent perception and control technology, this paper systematically reviews the development of visual and non-visual perception technology and discusses the application of multi-modal perception in complex environments. The latest development of motion control and force control technology is analyzed, and its importance in fine operation is expounded. The future development direction of robot intelligent perception and control technology is proposed, and the necessity of enhancing real-time robustness and system integration is emphasized, which provides theoretical support and technical reference for promoting the continuous innovation of robot intelligent perception and control technology.

Normative Data for Motor Free Visual Perception Test- 4 in Indian Population Between 50-60 Years in Normal Healthy Individuals

Visual perceptual skills are critical for performing various tasks, making them a key focus in Occupational Therapy, particularly across different diagnoses. The Motor-Free Visual Perception Test (MVPT-4) is a standardized tool used to evaluate visual perception independent of motor abilities in individuals aged 4 to 80+. Although MVPT-4 has been widely used across diverse populations, normative data for adults, especially in the Indian population, is limited. This study aims to establish normative data for healthy North Indian adults aged 50-60. A total of 100 individuals (50 males, 50 females) were assessed using MVPT-4. Test scores were recorded, and statistical analysis was performed. Results indicated an average score of 27.2 (SD = 7.31) in females and 29.32 (SD = 7.45) in males. These findings contribute to a deeper understanding of visual perceptual abilities in the Indian population and provide normative data for future research and clinical practice in the region.

Characteristic of temporal light modulation reference meter and standard light source for calibration and verification

Abstract The measurement of the temporal light modulation (TLM) of a light source is gaining increasing importance owing to its detrimental effects on human vision and health. To establish traceability in TLM measurements, the National Institute of Metrology, China (NIM), established a TLM reference meter (TLMM) and standard TLM light source (STLS) for calibration and verification services. The TLMM and STLS employ detector- and source-based approaches, respectively, to define the standards for TLM quantities. The TLMM rapidly acquires light waveform data from the TLM light sources and calculates corresponding TLM quantities. Frequency calibration result is with an expanded uncertainty of 1 × 10−5, frequency response function is almost flat in range of 1 kHz–25 kHz, and response time is less than 8 × 10−6 s. The electrical noise was evaluated at approximately 1000:1–5000:1. The STLS generates an arbitrary TLM waveform by converting the virtual waveform into an actual LED output light waveform. For the frequency response function, the relative amplitude decreased by 1% with a frequency of 15 kHz, and response time was less than 5 × 10−6 s. The linearity of the frequency dependency was studied with ratio varied by approximately 0.19% when the periodic triangle excitation in frequency band from 1 Hz to 1 kHz. Experimental comparisons demonstrated that TLM quantities (frequency, percentage flicker, flicker index, short-term flicker severity P st LM , and stroboscopic effect visibility measure ( SVM )) are consistent between the TLMM and STLS devices on 14 typical TLM waveforms. For P st LM , most of the deviations were in the range of −0.019–0.079 (except for one with 0.182). For SVM , most of the deviations were in the range of −0.001–0.003 (except for one with 0.014).

Supplemental Material for Employee Suicide: The Effect of Cause, Employer Response, and Gender on Perceptions of Organisational Brand Image and Support

Supplemental Material for Employee Suicide: The Effect of Cause, Employer Response, and Gender on Perceptions of Organisational Brand Image and Support

Does patient-ventilator asynchrony really matter?

Past observational studies have reported the association between patient-ventilator asynchronies and poor clinical outcomes, namely longer duration of mechanical ventilation and higher mortality. But causality has remained undetermined. During the era of lung and diaphragm protective ventilation, should we revolutionize our clinical practice to detect and treat dyssynchrony? Clinicians' ability to recognize asynchronies is typically low. Automatized softwares based on artificial intelligence have been trained to largely outperform human eyesight and are close to be implemented at the bedside. There is growing evidence that in susceptible patients, dyssynchrony may lead to ventilation-induced lung injury (or patient self-inflicted lung injury) and that clusters of such dyssynchronous events have the highest association with poor outcomes. Dyssynchrony may also be associated with harm indirectly when it reflects over-assistance or over-sedation. However, the occurrence of reverse triggering by means of low inspiratory efforts during passive ventilation may prevent diaphragm dysfunction and atrophy and be beneficial. Most recent evidence on the topic suggests that synchrony between the patient and the mechanical ventilator is a critical element for protecting lung and diaphragm during the time of invasive mechanical ventilation or may reflect inadequate settings or sedation. Therefore, it is a complex situation, and clinical trials are still needed to test the effectiveness of keeping patient-ventilator interaction synchronous on clinical outcomes.

Social Media Use and Regional Image Perception: An Empirical Study Based on the Gansu Image Perception Questionnaire Survey

This study, using Gansu Province in China as a case, explores the impact of social media use on regional image perception. It finds that social media enhances users' perceptions of economic, political, cultural, and environmental images, indirectly increasing their satisfaction with the Gansu region. The research shows that social media plays a positive role in constructing regional images, particularly through user-generated content on short video platforms, which is more attractive and shareable. Among these, environmental and cultural images have the most significant impact on regional image compared to economic and political images. Therefore, it is recommended that regional governance bodies shift their promotional strategies to stimulate internet users' creative enthusiasm to improve the effectiveness of regional image dissemination.

Alignment of color discrimination in humans and image segmentation networks.

The experiments allowed by current machine learning models imply a revival of the debate on the causes of specific trends of human visual psychophysics. Machine learning facilitates the exploration of the effect of specific visual goals (such as image segmentation) by different neural architectures in different statistical environments in an unprecedented manner. In this way, (1) the principles behind psychophysical facts such as the non-Euclidean nature of human color discrimination and (2) the emergence of human-like behaviour in artificial systems can be explored under a new light. In this work, we show for the first time that the tolerance or invariance of image segmentation networks for natural images under changes of illuminant in the color space (a sort of insensitivity region around the white) is an ellipsoid oriented similarly to a (human) MacAdam ellipse. This striking similarity between an artificial system and human vision motivates a set of experiments checking the relevance of the statistical environment on the emergence of such insensitivity regions. Results suggest, that in this case, the statistics of the environment may be more relevant than the architecture selected to perform the image segmentation.

Virtual cleaning of sooty mural hyperspectral images using the LIME model and improved dark channel prior

Murals, as important carriers of cultural heritage and historical records, showcase artistic, aesthetic, social, and political significance. In ancient times, religious activities such as burning incense and candles in temples led to many murals being polluted by soot, causing them to darken, lose details, and, in severe cases, completely blacken. As a result, the development of efficient virtual cleaning methods has become a key strategy for addressing this issue. In this study, we use synthetic true colour and false colour images in different bands of the hyperspectral spectrum, and use a guided filter fusion technique to fuse these two images into a new image of the sooty mural. Through analyzing the histograms and colour distribution scatterplots of the synthetic sooty mural images, we observed significant similarities to low-luminance images. To enhance the synthesized murals, we applied the LIME model. In addition, comparisons of the histograms and colour distribution scatterplots of the enhanced sooty mural images with those of haze images revealed notable similarities. Therefore, we applied the dark channel prior algorithm to remove soot from the mural images. Considering that soot particles are larger than haze particles, we introduced guided filtering to refine the transmission map and created a nonlinear transformation function to enhance its details. In terms of both visual perception and quantitative analysis, the proposed method significantly outperforms previous methods in the virtual cleaning of sooty murals. This technology can not only restore the colours and details of murals but also provide new clues for subsequent mural studies, allowing people to once again appreciate the true beauty of the murals.

Measuring Street Quality: A Human-Centered Exploration Based on Multi-Sourced Data and Classical Urban Design Theories

Advancements in analytical tools have facilitated numerous studies on perceived street quality. However, most have focused on limited aspects of street quality, failing to capture a comprehensive perception. This study introduces a quantitative approach to holistically measure street quality by integrating three key dimensions: visual perception, network accessibility, and functional diversity. Using Beijing and Shanghai as case studies, we employed artificial neural networks to analyze street view images and quantify the visual characteristics of streets. Additionally, street network accessibility was assessed through spatial design network analysis, and functional diversity was evaluated using the entropy of points of interest (POIs) data. The evaluation results were combined using the analytic hierarchy process. The reliability and accuracy of this method were validated through further testing. Our approach offers a human-centered, large-scale measurement framework, providing valuable insights for urban street renewal and design.