Image Formation Research Articles

Understanding The Impact of Culture as A Destination Image Attribute on Sustainable Tourism

This study aims to underscore the pivotal role of cultural attributes in shaping destination image and examines their implications for sustainable tourism. Through a comprehensive review of relevant literature, the study proposes a conceptual framework to delve into the multifaceted effects of cultural attributes on destination image formation and offers strategies and guidelines for fostering sustainable tourism practices. The findings of the study underscore the significance of cultural attributes in bolstering the competitiveness of destination images in sustainable tourism, while also analyzing their interconnectedness across social, economic, and environmental dimensions. The societal impact of this study lies in highlighting the pivotal role of cultural attributes in shaping perceptions and allure associated with destination images, thereby furnishing a vital reference point for crafting and executing a targeted research agenda in urban tourism. Ultimately, this study furnishes a valuable theoretical framework and perspective conducive to the advancement of sustainable tourism endeavors.

Preliminary Study Based on Myocardial Infarction Classification of 12-Lead Electrocardiography Images with Deep Learning Methods

In contemporary medicine, the development of computer-aided diagnostic systems using Electrocardiography (ECG) signals has gained significance for the diagnosis of heart diseases. Myocardial infarction (MI) is recognized as the condition where blood flow to the heart muscle is obstructed due to blockages in coronary vessels. In this study, four deep learning approaches were employed to automatically identify different MI conditions (STEMI, NSTEMI, USAP) using images generated from 12-lead ECG signals. The utilized architectures include deep neural networks such as Visual Geometry Group-16 (VGG-16), AlexNet, Residual Neural Network (ResNet), SqueezeNet and an ensemble model composed of these networks. With the proposed method, classification was performed based on 10-second grayscale images of 12-lead ECG signals for HC-STEMI, HC-NSTEMI, HC-USAP, and NSTEMI-STEMI conditions. According to the obtained results, the HC-STEMI group achieved the highest performance with a cross-validated 0.8237 F1 score using the AlexNet architecture. Among the novel contributions of this study is the image-based ECG classification method that can be more easily adapted to clinical applications and the analysis of the potential use of detecting different MI conditions in clinical practices. In conclusion, this study sheds light on future research by demonstrating the significant potential of using multi-channel ECG signals in image format for MI diagnosis, paving the way for advancements in this field.

Туристік бизнесті дамытудағы ақпараттық жүйелер, интернет-платформалардың алатын орны

Digitalization of all sectors, especially economic sectors, in the country makes it possible to take leading positions on the world stage, as well as improve the quality of life of citizens and create favorable conditions for the development of entrepreneurship. A dynamically developing tourism business with large incomes in the economic sphere is also engaged in digitalization. The mass introduction of digital technologies will give impetus to the development of traditional basic industries by improving the quality of services, increasing their competitiveness, including on the international market. Thus, as a result of digitalization, the provision of top-level services in the tourism sector will be ensured, which, in turn, will lead to a positive formation of the country's tourist image and capital growth. Digitalization of the tourism business optimizes the problem, although it does not solve all the problems. For example, in the field of tourism, if information is provided without correct statistical calculations, then state support measures will also not be directly adjusted. And with the help of digitization, many adjustments and deviations in information can be avoided.

Brainwaves and Emotional Responses to Changes in Red Eye Shadow Brightness and Saturation

Purpose: Investigating how Munsell’s primary color (i.e., red eye shadow) influences human brain and emotional responses to establish foundational data for image formation. Methods: We analyzed the relative power differences in background brainwaves due to variations in five levels of brightness and saturation for each eye both in male and female individuals in their twenties. We used the coefficient of variation for brainwaves for our analysis. In addition, we assessed the preferences for altered red eye shadow and the emotional responses to determine the average preference index. We performed statistical analysis for brainwaves and emotional evaluations using repeated measure analysis of variance in SPSS, followed by post hoc testing with Tukey’s test at a 95% confidence level. Results: The brainwave and emotional responses obtained by presenting red eye shadow brightness and saturation modification-created stimuli revealed that both men and women exhibited beta and gamma wave activation and alpha wave inactivation upon low brightness and high saturation stimuli. Men were more sensitive to brightness stimuli, whereas women displayed higher-level variability in brainwave responses upon saturation changes. Both genders preferred red eye shadows with brightness and saturation levels of 8 and 2, respectively. The perceptual image responses indicated that stimuli with low or high brightness were perceived as modern and subtle or clear and cute, respectively. Medium or high saturation stimuli were perceived as clear and natural or associated with a modern image, respectively, by both genders. Conclusion: This study demonstrated the impact of various color stimuli on the human brain and emotions. We consider the obtained results as foundational data, which could be used in future image consulting or when crafting desired image presentations.

Simplifying vein detection for intravenous procedures: A comparative assessment through near‐infrared imaging system

AbstractThe intravenous (IV) injection procedure can be a challenging task, especially for individuals with thin veins, obesity, or patients with damaged and pigmented skin. Therefore, the IV procedure necessitates a portable medical device that can be used for academic demonstrations to train medical students or by health care professionals to perform venipuncture. Vein visualization with a vein detector is principally based on the interaction of blood components with wavelengths of the electromagnetic spectrum (EMS). In this paper, we first present the process of image formation in the spectral band of the near‐infrared region (NIR) of EMS. Then, we introduce the image acquisition system with image processing to extract the veins as a noninvasive vein detection method. A Raspberry Pi (Model 4B), along with a night vision camera, serves as an image acquisition tool to capture skin area illuminated by NIR. Following this, the data is transferred to the laptop where it can be filtered and processed using Python image processing tools before being viewed on the monitor. The results achieved through the device are quite encouraging, as the image recognition between veins and adjacent tissues from the skin sample can be clearly marked. The functionality, accuracy, and simplicity associated with this vein detection system make it a potential device for IV placement and the morphological study of disease detection.

Learn More Manchu Words with A New Visual-Language Framework

Manchu language, a minority language of China, is of significant historical and research value. An increasing number of Manchu documents are digitized into image format for better preservation and study. Recently, many researchers focused on identifying Manchu words in digitized documents. In previous approaches, a variety of Manchu words are recognized based on visual cues. However, we notice that visual-based approaches have some obvious drawbacks. On one hand, it is difficult to distinguish between similar and distorted letters. On the other hand, portions of letters obscured by breakage and stains are hard to identify. To cope with these two challenges, we propose a visual-language framework, namely the Visual-Language framework for Manchu word Recognition (VLMR), which fuses visual and semantic information to accurately recognize Manchu words. Whenever visual information is not available, the language model can automatically associate the semantics of words. The performance of our method is further enhanced by introducing a self-knowledge distillation network. In addition, we created a new handwritten Manchu word dataset named (HMW), which contains 6,721 handwritten Manchu words. The novel approach is evaluated on WMW and HMW. The experiments show that our proposed method achieves state-of-the-art performance on both datasets.

Evolution of the Concept of <i>Barbarism</i> within the Context of the Franco-German Confrontation in the 1870s–1910s

Intolerance, demonization and hate speech used to be an integral part of Franco-German relations in the 19th century. The paper is devoted to the analysis of the image of barbarians within the context of the confrontation between France and Germany in 1870s–1910s. Creation and transformation of political myths is viewed through the lens of cultural artefacts. The study is topical because it allows for the analysis of the universal development of cultural confrontation on a specific example. The prime goal of the article is to identify the patterns of mutual stigmatization and account for their seeming anomalies or exceptions to these patterns: the German simultaneously viewed as well-educated and barbarian; common origins of barbarian and classical antique cultures resulting in opposing identities, etc. The aims of the study include: 1) to systematize empirical sources; 2) to assess the impact of historical and cultural context on the formation of the conceptual images of barbaric and civilized; 3) to single out basic dichotomies of the barbaric vs civilized opposition. The study was based on various cultural texts of the period, ranging from speeches, essays, and memoirs to posters and monuments. Methods used included comparative, semiotic, and imagological approaches, corpus, discourse, and content analyses. French newspaper clichés are compared to the Europe-wide opinions about Prussian militarism. As it turns out, the German way of thinking about the heroes of the past contributed to political stigmatization of Germany. Another finding, based on French literature, is the unobvious correlation between decadent world perception and political turmoil in France. Modernist poetry, literature and journalism compared the fall of the French Empire to the collapse of the Roman Empire and thus contributed to the conceptualizing of the German as barbarians. At the same time German propaganda attempted to defend itself by finding barbarians on the east and ironically quoting French sources to speak about actual war crimes. The results of the study show that seemingly chaotic hate speech and confrontation are determined on the basis of the specific rules related to age-old non-political traditions and current political circumstances marking the dichotomy between oneself and the other, friend or foe.

Spherical Pseudo-Cylindrical Representation for Omnidirectional Image Super-resolution

Omnidirectional images have attracted significant attention in recent years due to the rapid development of virtual reality technologies. Equirectangular projection (ERP), a naive form to store and transfer omnidirectional images, however, is challenging for existing two-dimensional (2D) image super-resolution (SR) methods due to its inhomogeneous distributed sampling density and distortion across latitude. In this paper, we make one of the first attempts to design a spherical pseudo-cylindrical representation, which not only allows pixels at different latitudes to adaptively adopt the best distinct sampling density but also is model-agnostic to most off-the-shelf SR methods, enhancing their performances. Specifically, we start by upsampling each latitude of the input ERP image and design a computationally tractable optimization algorithm to adaptively obtain a (sub)-optimal sampling density for each latitude of the ERP image. Addressing the distortion of ERP, we introduce a new viewport-based training loss based on the original 3D sphere format of the omnidirectional image, which inherently lacks distortion. Finally, we present a simple yet effective recursive progressive omnidirectional SR network to showcase the feasibility of our idea. The experimental results on public datasets demonstrate the effectiveness of the proposed method as well as the consistently superior performance of our method over most state-of-the-art methods both quantitatively and qualitatively.

Abstract 3532: Transforming genomic data into images for enhanced deep learning in precision oncology

Abstract Purpose: In the realm of cancer treatment and research, the analysis of genomic data stands as a cornerstone for diagnosis and therapy. Recognizing this, our study embarks on an ambitious journey to revolutionize how we interpret such data. By converting complex genomic data into a two-dimensional (2D) image format leveraging feature correlations, we pave the way for more profound insights and applications in clinical decision-making. This innovative approach utilizes 2D convolutional neural networks (CNNs) to analyze these structured images, promising a leap forward in the accuracy and utility of such data. Methodology: The core of our method involves a novel transformation of tabular genomic data into an image format. We start with m patient samples, each characterized by n genes, and calculate a pairwise correlation matrix to capture the intricate relationships between these genes. Concurrently, we generate an Euclidean distance matrix that represents the distances between n points in a 2D grid. The next step involves the optimization of the Gromov-Wasserstein discrepancy to align these two matrices, resulting in a transformation matrix, T. When applied to the tabular data, T transforms it into m distinct, informative images. These images are then processed using a 10-layer CNN, comprising a convolutional layer (3 × 3 kernel), three dense layers, two relu layers, and a dropout layer. Data Sources: We created images, termed 'genomaps', from genomic data collected from three diverse patient groups: 130 individuals (cancerous and normal) from Stanford Hospital, 230 breast cancer patients from The Cancer Genome Atlas (TCGA), and a larger group of 1572 patients from the Memorial Sloan Kettering (MSK) Cancer Center. The created genomaps are visually interpretable, offering clear differentiation between cancerous and non-cancerous samples, which is pivotal for accurate analysis. Outcomes: Our method demonstrates a remarkable 8% improvement in survival prediction accuracy in comparison to the existing methods. The employment of DeepSHAP in our analysis has allowed us to identify critical genes with greater precision than conventional methods. The p53 gene family, notably TP53 and TP63, was identified as significantly mutated in over 50% of all cancer types analyzed. Another critical discovery was the ERBB gene family, encompassing EGFR, ERBB2, ERBB3, ERBB4, which plays a dual role in tumor proliferation and influencing the immune response against tumors, a factor critical for immunotherapy strategies. Conclusion: This research introduces a transformative approach to represent genomic data, utilizing 2D CNNs for in-depth analysis. It sets a new benchmark in classification and regression accuracy, offering a more interpretable pathway for biomarker discovery, significantly contributing to the field of personalized medicine and cancer research. Citation Format: Md Tauhidul Islam, Lei Xing. Transforming genomic data into images for enhanced deep learning in precision oncology [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 3532.

Abstract 4933: A workflow for cloud-based AI development of multiplex IF image analysis using the OMERO Plus platform

Abstract Background: The advancement of AI in digital pathology is essential for analyzing image data containing billions of cells. This is true for digital pathology workflows using bright-field images, but this is even more critical for multiplex immunofluorescence (mIF) whole-slides due to the increased number of fluorescent channels and bit-depth per pixel. These complexities necessitate improved data management, and efficient visualization for both the mIF image and AI-based analysis results. We present a workflow using OMERO Plus, AWS SageMaker, and Ultivue software for AI model development in mIF. Method: Images were uploaded to AWS S3 and imported to the data management system OMERO Plus, where they were organized into training and validation groups. Annotations: PathViewer was used to place rectangular fields of view (FOV) in locations selected by the pathologist where all positive cells for each target marker were manually annotated. Data Preparation: Using the OMERO Plus API, annotation coordinates, image tiles, and image metadata were extracted and formatted for model training. All training data is under a version control system using Gitlab and Data Version Control tools using AWS S3. Additionally, strict quality checks were performed to ensure data integrity. Training: For both semantic segmentation and cell object detection, the training data was used in a customized training workflow using AWS SageMaker, including modern deep learning architectures, data augmentation, training logging, evaluation, and deployment. All AI models were trained using Pytorch and deployed with the open model format ONNX. Results Visualization: AI segmentation results of cells and regions are visualized using PathViewer as overlays on the original images by converting them to the OME-NGFF label image format that uses Zarr multidimensional arrays. Results: We utilized 58 and 24 mIF images for training and validation, respectively, on 15 markers using ISP Ultivue technology with 1681 and 296 FOVs, encompassing over 60K cells. Using this data and methodology we trained 3 semantic segmentation models and over five object detection models suitable for all twenty-four OmniVUE markers. For instance, the CD3 positive cell detection model, trained with 390 FOVs, achieved highly accurate results: f1-score of 0.915 evaluated on 34 FOVs. Conclusions: The combination of OMERO Plus, PathViewer, AWS SageMaker, and internal tools can successfully address mIF Big Data data management complexities in a cloud environment for digital pathology. The architecture is highly scalable, cost effective, and reliable, enabling a very efficient AI workflow. Citation Format: Ruben Cardenes, Erin Diel, Douglas Wood, Je Lee, Angela Vasaturo, Martin Schulze, Lorenz Rognoni. A workflow for cloud-based AI development of multiplex IF image analysis using the OMERO Plus platform [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 4933.

Image Analysis and Classification Using HRSVM-CNN for Land-Cover Classification by Using Remote Sensing Images

Objective: To effectively analyze diverse satellite images and derive valuable insights, it's crucial to employ efficient methods for classification and image processing. However, due to imperfections in image formats and sensor data, satellite imagery often contains flaws and inaccuracies, requiring extra steps to enhance its quality. The proposed solution involves two key techniques: segmenting the input image using a Hybrid HRSVM-CNN and classifying the resulting high-resolution remote sensing image using a Convolutional Neural Network. This combined approach addresses the challenges posed by image inconsistencies and aims to improve the accuracy and efficiency of current methods for satellite image analysis. Methods: In this research, a high-resolution Support Vector Machine-Convolutional Neural Network (Hybrid HRSVM-CNN) and texture characteristics are used to create an automated land identification method for satellite Remote sensing (RS) images. This approach's main focus is segmentation using the Bendlet Transform and Improved Chan-Vese, and it also does classification using a Hybrid HRSVM-CNN based on feature extraction and gray-level co-occurrence matrix algorithm. Findings: The proposed classification method's accuracy was evaluated against several other classification algorithms, including Semi-Supervised Graph Based Method (SSG), Conditional Random Fields (CRF), k-Nearest Neighbor (KNN), and Bi-layer Graph-based Learning (BLGL), Support Vector Machine (SVM), and Artificial Neural Networks (ANN). When compared with existing methods, the findings of the proposed method display excellent accuracy of 98.83%. Novelty: To start, an adaptive median filter is used to pre-process the satellite Remote sensing images, removing unwanted noises and other impacts. Following pre-processing, the image is segmented using the Bendlet Transform and Improved Chan-Vese algorithms. Gray-level co-occurrence matrix is utilized to extract texture information, and the Hybrid HRSVM-CNN is then used to categorize the various types of land. These applications frequently have a variety of issues that affect the categorization accuracy. A few significant factors, like location, irregularity, form, and diameter, reduce the process's overall accuracy. This research focuses on presenting a unique Land-Cover classification model in order to address such problems. The UC Merced Land Use dataset was considered in this research. Keywords: Bendlet Transform and Improved Chan­Vese Segmentation, Hybrid HRSVM­CNN, Land detection system, Satellite remote sensing images, Adaptive median filter, Gray­ level co­occurrence matrix

Nature-Based Destination Image: an empirically validated multidimensional measurement model in the Brazilian Amazon destination.

This paper aimed to empirically validate a multidimensional measurement model for the Nature-Based Tourism Destination Image - NBDI, including the cognitive, affective, and conative components, present in the construct of variables tested by Chaves and Anjos, (2022). The validation of the NBDI measurement model relied on a non-probability quota-based sample consisting of 592 international tourists who visited the Legal Amazon destination. The results pointed out that the NBDI can be measured by a three-dimensional construct of 3rd Order, where the image components presented a positive influence in the formation of the overall image, being the affective component of greater influence, followed by the cognitive and conative, presenting convergent and discriminant validity. The study contributes to the theoretical discussions about NBDI, offering a multidimensional scale with reliable psychometric property and validated with 57 variables, scarce subjects in the scientific literature, as well as, in the marketing strategies of public and private institutions.

«Ландшафт образов»: символические репрезентации Якутии в геопоэтике политических ссыльных

In this article, based on the historical and cognitive analysis of the artistic heritage of political exiles (XIX–XX centuries), who influenced the formation of intellectual history and culture of Yakutia, symbolic images associ-ated with the North as the “periphery” of the Russian Empire are revealed. In the mental mapping of artistic space, positive and negative images constructed on the basis of the “supporting metaphor” – Cold – are high-lighted. It is concluded that this concept of Cold influenced the formation of the main metaphorical image of Yakutia as the “edge of the earth” and created a layer of various concepts, knowledge, associations and experi-ences consisting of conceptual and value-symbolic components. It has been revealed that depending on the emotional and sensory perception of the surrounding space, one can observe the transformation of the mental image of the North. The collective image of the North includes many elements that allow us to present it as a mythologeme, and on the other hand, to consider the variations of the mentality of an exile forced to be in a completely alien world. Thus, despite the fact that imperial policy actively promoted the image of the North as an exotic and rich land, the “narrative of captivity” imposed on the works of political exiles transmits the image of the space of exile associated with the North as a “land of suffering” and “trials”.

Visual poetics of the story by Grigory Baklanov "South of the main offensive"

The subject of research in this article are the characteristics and specificity of visual images and codes in the work of Grigory Baklanov "South of the main offensive". The relevance of the work lies in the fact that the scientific works devoted to the prose of Baklanov are represented by few works that study the poetics of psychologism in the writer's prose (A. E. Abramova) or the military prose of the Sixtiers as a whole (for example, in the study of T. G. Bogoradova and I. L. Startseva, Baklanov's prose is presented only as a part of the aggregate object). The writer's work is more widely discussed in the critical works of L. I. Lazarev, I. A. Dedkov and L. V. Oborin. Despite the importance of the figure of Grigory Baklanov in the "lieutenant prose", the sensual features of his prose are practically not studied. Meanwhile, Grigory Baklanov was one of the leading authors of "lieutenant prose". Thus, V. V. Bykov noted the primary influence on his work of the story "South of the main offensive" (1957) – the study of the visual poetics of this very work seems relevant. The methodological basis of the research was the immanent analysis, the comparative-comparative method, as well as B. A. Uspensky's approach to the analysis of the "point of view" in fiction. Aim of the research is to study the visual poetics of the story "South of the main offensive". The aim of the study is to consider the problem of visual poetics in modern literary studies, as well as to determine the functions and role of visual poetics in the artistic text of the story "South of the main offensive". In the work of Grigory Baklanov, the aspect of sensory poetics, which consists in appealing to the reader's perceptual experience through the text, is particularly pronounced – it is this aspect that is responsible for the formation of the sensory image of the artistic world of the work. In this study, visuality is considered as the main aspect of the reader's perception.

Image analysis and resolution for detection-based synthetic-aperture passive source localization

This paper follows a detection-theoretic approach for using synthetic-aperture measurements, made at multiple moving passive receivers, in order to form an image showing the locations of stationary sources that are radiating unknown electromagnetic or acoustic waves. The paper starts with a physics-based model for the propagating fields, and, following the general approach of McWhorter et al (2023 arXiv:2302.06816, IEEE Open J. Signal Process. 4 437–51), derives a detection statistic that is used for the image formation. This detection statistic is a quadratic function of the data. Each point in the scene is tested as a possible hypothesized location for a source, and the detection statistic is plotted as a function of location. Because this image formation process is nonlinear, the standard linear methods for determining resolution cannot be applied. This paper shows how to analyze the detection image by first writing the noiseless image as a coherent sum of shifted complex ambiguity functions of the source waveform. The paper then develops a technique for calculating image resolution; resolution is found to depend on the sensor-source geometry and also on the properties (bandwidth and temporal duration) of the source waveform. Optimal filtering of the image is given, but a simple example suggests that optimal filtering may have little effect. Analysis is also given for the case in which multiple sources are present.

Strengthening the institution of the family by social work methods

The article considers the possibilities of preparing teenagers for family life through the use of active-forming technologies of social work. The crisis of the institution of the family, the loss of their functions by the family has attracted the attention of scientists for more than a decade. The crisis of the family is closely connected with the spread of negative processes in society as a whole. Among them: an increase in the number of teenage pregnancies, the incidence of HIV, a decrease in the birth rate, etc. This confirms the relevance of work in the field of sexual education of schoolchildren, the formation of a positive image of the family in society, and increasing the parental and marital competence of young people. Despite this, classes to prepare schoolchildren for family life in Russian educational institutions are still not held. The authors propose to use active-shaping technologies of social work for these purposes. The interactive nature of the proposed course has many advantages. It allows teenagers to be more open in discussing intimate topics, ask clarifying questions and express their own opinions during class. The article discusses the psychosocial characteristics of adolescents, presents the results of testing an active-forming program for preparing adolescents for family life in work with high school students.

A Regional 100 m Soil Grid-Based Geographic Decision Support System to Support the Planning of New Sustainable Vineyards

A WebGis tool called GoProsit has been developed to support winegrowers in planning a new sustainable vineyard and in the identification of high-quality terroir in Tuscany, Central Italy, by providing various information on soils, climate, hydrological risks, and fertilization. GoProsit, hosted by the web platform GEAPP, is a free, user-friendly, and interactive Geographic Decision Support System (GDSS). Soil data behind the WebGis tool has a 1 ha resolution, achieved by processing the legacy vector-type soil database of the Tuscany Region with the DSMART (Disaggregation and Harmonization of Soil Map Units Through Resampled Classification Trees as supervised classification) algorithm, which disaggregated the map to 297,023 vineyard grid cells. Each grid cell holds climatic and pedologic information, along with physical and chemical features for each horizon of the most probable soil. GoProsit also provides soil maps in image format obtained by georeferencing about 50 historical soil maps (1969–2012). Finally, GoProsit runs and returns the outputs of six models: (a) carbon footprint, (b) potential erosion and maximum vine row length compatible with tolerable erosion, (c) potential water stress, (d) risk of runoff/waterlogging, (e) identification of suitable rootstocks, and (f) nutritional needs before planting. Statistics of the main model results for the investigated area are reported. This promising tool will soon be usable for the whole Italian territory; however, its potential makes it suitable for use in any wine-growing district.

Three-Dimensional Reconstruction Pre-Training as a Prior to Improve Robustness to Adversarial Attacks and Spurious Correlation.

Ensuring robustness of image classifiers against adversarial attacks and spurious correlation has been challenging. One of the most effective methods for adversarial robustness is a type of data augmentation that uses adversarial examples during training. Here, inspired by computational models of human vision, we explore a synthesis of this approach by leveraging a structured prior over image formation: the 3D geometry of objects and how it projects to images. We combine adversarial training with a weight initialization that implicitly encodes such a prior about 3D objects via 3D reconstruction pre-training. We evaluate our approach using two different datasets and compare it to alternative pre-training protocols that do not encode a prior about 3D shape. To systematically explore the effect of 3D pre-training, we introduce a novel dataset called Geon3D, which consists of simple shapes that nevertheless capture variation in multiple distinct dimensions of geometry. We find that while 3D reconstruction pre-training does not improve robustness for the simplest dataset setting, we consider (Geon3D on a clean background) that it improves upon adversarial training in more realistic (Geon3D with textured background and ShapeNet) conditions. We also find that 3D pre-training coupled with adversarial training improves the robustness to spurious correlations between shape and background textures. Furthermore, we show that the benefit of using 3D-based pre-training outperforms 2D-based pre-training on ShapeNet. We hope that these results encourage further investigation of the benefits of structured, 3D-based models of vision for adversarial robustness.

Beyond pixels and ciphers: Navigating the advancements and challenges in visual cryptography

<p>Visual cryptography (VC) has emerged as a pivotal solution for secure information transmission, leveraging its unique capability to encrypt images in a user-friendly and accessible manner. This survey paper provides an in-depth analysis of various VC methods, highlighting their distinct encryption and decryption techniques, applicability, and security levels. The study delves into the technical specifications of each VC type, offering insights into secret image formats, the number of secret images used, types of shares, pixel expansion, and complexity. Significant attention is given to the practical applications of VC, ranging from secure document verification and anti-counterfeiting measures to digital watermarking and online data protection. The paper also identifies key challenges in the field, such as image quality retention post-decryption, computational efficiency, and scalability. Future prospects of VC are explored, particularly its potential integration with emerging technologies like AI and blockchain. This survey aims to provide a comprehensive understanding of VC’s current state, its diverse applications, and the future possibilities, making it a valuable resource for researchers and practitioners in the field of data security and cryptography.</p>

Deep learning of movement behavior profiles and their association with markers of cardiometabolic health

BackgroundTraditionally, existing studies assessing the health associations of accelerometer-measured movement behaviors have been performed with few averaged values, mainly representing the duration of physical activities and sedentary behaviors. Such averaged values cannot naturally capture the complex interplay between the duration, timing, and patterns of accumulation of movement behaviors, that altogether may be codependently related to health outcomes in adults. In this study, we introduce a novel approach to visually represent recorded movement behaviors as images using original accelerometer outputs. Subsequently, we utilize these images for cluster analysis employing deep convolutional autoencoders.MethodsOur method involves converting minute-by-minute accelerometer outputs (activity counts) into a 2D image format, capturing the entire spectrum of movement behaviors performed by each participant. By utilizing convolutional autoencoders, we enable the learning of these image-based representations. Subsequently, we apply the K-means algorithm to cluster these learned representations. We used data from 1812 adult (20–65 years) participants in the National Health and Nutrition Examination Survey (NHANES, 2003–2006 cycles) study who worn a hip-worn accelerometer for 7 seven consecutive days and provided valid accelerometer data.ResultsDeep convolutional autoencoders were able to learn the image representation, encompassing the entire spectrum of movement behaviors. The images were encoded into 32 latent variables, and cluster analysis based on these learned representations for the movement behavior images resulted in the identification of four distinct movement behavior profiles characterized by varying levels, timing, and patterns of accumulation of movement behaviors. After adjusting for potential covariates, the movement behavior profile characterized as “Early-morning movers” and the profile characterized as “Highest activity” both had lower levels of insulin (P < 0.01 for both), triglycerides (P < 0.05 and P < 0.01, respectively), HOMA-IR (P < 0.01 for both), and plasma glucose (P < 0.05 and P < 0.1, respectively) compared to the “Lowest activity” profile. No significant differences were observed for the “Least sedentary movers” profile compared to the “Lowest activity” profile.ConclusionsDeep learning of movement behavior profiles revealed that, in addition to duration and patterns of movement behaviors, the timing of physical activity may also be crucial for gaining additional health benefits.