Style Transformation Research Articles

Data Augmentation for Semantic Segmentation Using a Real Image Dataset Captured Around the Tsukuba City Hall

We are exploring the use of semantic scene understanding in autonomous navigation for the Tsukuba Challenge. However, manually creating a comprehensive dataset that covers various outdoor scenes with time and weather variations to ensure high accuracy in semantic segmentation is onerous. Therefore, we propose modifications to the model and backbone of semantic segmentation, along with data augmentation techniques. The data augmentation techniques, including the addition of virtual shadows, histogram matching, and style transformations, aim to improve the representation of variations in shadow presence and color tones. In our evaluation using images from the Tsukuba Challenge course, we achieved the highest accuracy by switching the model to PSPNet and changing the backbone to ResNeXt. Furthermore, the adaptation of shadow and histogram proved effective for critical classes in robot navigation, such as road, sidewalk, and terrain. In particular, the combination of histogram matching and shadow application demonstrated effectiveness for data not included in the base training dataset.

Artistic Style Transformation Based on Generative Confrontation Network

Artistic Style Transformation Based on Generative Confrontation Network

Research on the Underlying Principles and Deep Learning Algorithms based on Image Style Conversion Techniques

Image style transformation techniques are an important research area in the field of computer vision, aiming to combine the content of one image with the artistic style of another image to create new images with the content of the input image and the style of the artistic style image. The development of this field has benefited from the rapid development of deep learning algorithms, especially the application of techniques such as Convolutional Neural Networks (CNN) and Generative Adversarial Networks (GAN). At the heart of image style transformation is how to capture both the content and artistic style of the input image. This is typically achieved by using different levels of activation of the convolutional neural network as content and style representations. The content representation captures the object and structural information in the image, while the style representation captures the texture and colour information of the image. To capture the style of an image, a Gram matrix is typically used to measure the correlation between features at a particular level. This matrix representation allows us to capture texture information and colour distribution, thus enabling style migration. The goal of the image style transformation model is to minimise both content loss and style loss to produce synthetic images. Content loss typically uses Euclidean distance to measure the difference between content representations, while style loss involves the distance between Gram matrices. GANs have become an important tool in image style transformation. The generator network is responsible for generating synthetic images and the discriminator network evaluates the realism of the generated images. Through adversarial training, the generator and discriminator networks constantly compete to improve the quality of the generated images. Finally, this study highlights the potential applications of image style conversion techniques in the fields of art creation, image editing and virtual reality, and proposes directions for future research on deep learning algorithms and underlying principles to further improve the efficiency and quality of image style conversion techniques. The continuous development of this field will bring new opportunities for image processing and creative applications.

Customized meta-dataset for automatic classifier accuracy evaluation

Automatic classifier accuracy evaluation (ACAEval) on unlabeled test sets is critical for unseen real-world environments. The use of dataset-level regression on synthesized meta-datasets (comprised of many sample sets) has shown promising results for ACAEval. However, the existing meta-dataset for ACAEval is created using simple image transformations such as rotation and background substitution, which can make it difficult to ensure a reasonable distribution shift between the sample set and the test set. When the distribution shift is large, it becomes challenging to estimate the classifier accuracy on the test set using those sample sets. To ensure more robust ACAEval, this paper attempts to customize a meta-dataset in which each sample set has a reasonable distribution shift to the test set. An intra-class cycle-consistent adversarial learning (ICAL) method is introduced to transfer the style of a labeled training set to the style of the test set, by jointly considering the domain shift issue, the label flipping issue (the semantic information may be changed after style transformation), and the diversity of multiple sample sets in the meta-dataset. Experiments validate that under the same experimental setup, our method outperforms the existing ACAEval methods by a good margin, and achieves state-of-the-art performance on several standard benchmark datasets, including digit classification and natural image classification.

Ring artifacts correction for computed tomography image using unsupervised contrastive learning

Objective. Computed tomography (CT) is a widely employed imaging technology for disease detection. However, CT images often suffer from ring artifacts, which may result from hardware defects and other factors. These artifacts compromise image quality and impede diagnosis. To address this challenge, we propose a novel method based on dual contrast learning image style transformation network model (DCLGAN) that effectively eliminates ring artifacts from CT images while preserving texture details. Approach. Our method involves simulating ring artifacts on real CT data to generate the uncorrected CT (uCT) data and transforming them into strip artifacts. Subsequently, the DCLGAN synthetic network is applied in the polar coordinate system to remove the strip artifacts and generate a synthetic CT (sCT). We compare the uCT and sCT images to obtain a residual image, which is then filtered to extract the strip artifacts. An inverse polar transformation is performed to obtain the ring artifacts, which are subtracted from the original CT image to produce a corrected image. Main results. To validate the effectiveness of our approach, we tested it using real CT data, simulated data, and cone beam computed tomography images of the patient’s brain. The corrected CT images showed a reduction in mean absolute error by 12.36 Hounsfield units (HU), a decrease in root mean square error by 18.94 HU, an increase in peak signal-to-noise ratio by 3.53 decibels (dB), and an improvement in structural similarity index by 9.24%. Significance. These results demonstrate the efficacy of our method in eliminating ring artifacts and preserving image details, making it a valuable tool for CT imaging.

The Evolution of Chinese Cinema in the New Century: From Martial Arts Classics to New Mainstream Blockbusters

This study explores the evolution of the new century film from martial arts to the new mainstream film blockbuster era. Firstly, the article reviews the heyday of martial arts films and their influence on film industry and culture. Then, it studies the transformation and evolution of martial arts films in the new century and the background of the rise of new mainstream films in this period. Subsequently, the article analyzes the characteristics and value orientation of the new mainstream blockbuster, as well as the fusion and difference between martial arts and the new mainstream. Furthermore, the audience and market performance of the two film styles and the influence of technological progress on martial arts and new mainstream films are studied. Finally, the article discusses the roles and challenges of directors and producers in the transformation of film style. This research reveals the evolution and diversification of the film industry in the new century, as well as the interaction and development between different film genres.

Spectrum and Style Transformation Framework for Omni-Domain COVID-19 Diagnosis

Coronavirus disease 2019 (COVID-19) is an ongoing global pandemic and profoundly affects almost all people around the world. Thus, many automatic diagnosis methods based on computed tomography (CT) images have been proposed to reduce the workload of radiologists. Most of the existing methods focus on the in-domain predictions, i.e., the training and testing have similar distributions, which is impractical in real-world situations, since the CT images can be collected from different devices and in different hospitals. To improve the diagnosis performance of COVID-19 for both in-domain and out-of-domain data, this paper proposes a spectrum and style transformation framework for omni-domain COVID-19 diagnosis. To achieve this, we first present a spectrum transform module, which helps to discover the discriminating features of each domain to recognize the in-domain data. Then, we formulate a cross-domain stylization module, which learns the cross-domain knowledge to enhance the model generalization capability to deal with out-of-domain data. Moreover, our framework is a plug-and-play module that can be easily integrated into existing deep models. We evaluate our framework on four COVID-19 datasets and show our method consistently improves the diagnosis performance of various methods on both in-domain and out-of-domain data.

Style transformation super-resolution GAN for extremely small infrared target image

With the development of generative adversarial networks, the super-resolution technique of reconstructing a high-resolution image from a low-resolution has achieved excellent resolution results. However, small, low-resolution images are widespread, such as images taken by a thermal camera or with a lens far from the target. Extremely small target image super-resolution is a challenging problem. The main reason is that the small infrared target has fewer pixels and weaker features. The current optimization methods for the tiny target are mainly based on multi-scale feature fusion or super-resolution enhancement. The low-resolution images characterizing small targets are usually obtained by downsampling with high-resolution images during training, which is different from the style of the tiny target in actual detection applications, resulting in poor resolution. In order to solve the problem, we propose a new resolution network: Style Transformation Super-Resolution Generative Adversarial Network (STSRGAN). It contains two sub-networks: one is style transformation GAN to convert the style of the image, and the other is super-resolution GAN. STSRGAN transforms a blurry infrared small target into a clear target with a distribution similar to the training set. Then the resolution can be increased to get a better enhancement effect. The discriminator distinguishes whether the input comes from the generator or the actual image to assist in generating a better super-resolution image. Meanwhile, we produced an infrared Unmanned Aerial Vehicle (UAV) small target dataset with target pixels below 16 × 16. Our method proves better resolution enhancement of small IR targets and shows superior performance over other methods through experiments.

Music Quotation in Contemporary Non-Academic Art: Aspects of Genre and Style Transformation

"The problematic field of the article focuses on aspects of genre and style transformation of well-known musical content that is recognizable and stylistically significant. In contemporary non-academic art, the established semantics of the original source is subject to change under the influence of a new creative rethinking, the possibilities of which are only increasing in the context of the technologies of the twenty-first century. These possibilities are embodied in various visual formats: from direct citation to borrowing or processing with its variations. The relevance of this topic lies in the need to study the phenomenon of citation as such, as well as to research and systematize other means of genre and style transformation of musical material. The direct object of the study is mostly music of the so-called “light” genres, music of mass culture, due to more mobile, transparent and visual observation opportunities. The author offers a self-made classification of possible methods of borrowing someone else’s text as a means of compositional technique, as well as of individual genres: cover, remake, borrowing, allusion, sampling. In the context of the study of the phenomenon of “musical text”, the peculiarities of realization and perception of the artistic and integral result are traced. As a result, several types of citation that have become widespread in contemporary music are distinguished. Keywords: quote, citation, borrowing, autocitation, samples, cover, remake, collage, intertextuality. "

Prevalence and factors associated with overweight and obesity among adolescents aged 12-15 years in LongNan city，Gansu province，China

Background: With the rapid development of social and economic level, the rapid improvement of family economic income and the continuous transformation of social life style, the body mass index of adolescents and children has risen rapidly, the overweight and obesity rate has been significantly increased in social life, and it is now widely regarded as a serious global public health problem. Objectives：This present aimed to identify prevalence of Factors associated with Overweight and Obesity among 12-15 year old adolescents and children in Longnan City, Gansu Province" Methods：Nine schools in LongNan City were selected by questionnaire survey, and each school selected a class of students as survey objects to investigate the influencing factors such as height, weight, diet, exercise and family. Simple logistic regression model, multiple logistic regression model and backward elimination indicated were used for analysis. Results: A total of 620 adolescents and children were investigated in this study. Simple logistic regression model showed that gender, the number of fast food consumption per week, the number of chocolate consumption per week, the number of fruit consumption per week, the length of exercise per week, the number of exercises per week, the living area and family genetic history were correlated with the occurrence of obesity. Multiple logistic regression model and reverse elimination results showed that the risk of obesity was higher in males (OR=3.47) than in females. Statistically, eating fast food (OR=1.65), chocolate (OR=2.07), not participating in weekly exercise (OR=1.58), lack of weekly exercise time (OR=2.75), and not eating fruit (OR=1.53) were significantly associated with childhood obesity. We also found that children living in cities (OR=1.80) were more likely to be obese than those living in rural areas, and that children with a family genetic history of obesity (OR=2.20) were also at risk of developing obesity. Conclusion：In LongNan City, boys are more prone to obesity than girls. Eating behavior, number of exercises, length of exercise, area of residence and family genetic history are the main influencing factors for overweight and obesity in adolescents and children. Keywords: Adolescent Overweight，Childhood，Obesity.

Robust One-Shot Segmentation of Brain Tissues via Image-Aligned Style Transformation

One-shot segmentation of brain tissues is typically a dual-model iterative learning: a registration model (reg-model) warps a carefully-labeled atlas onto unlabeled images to initialize their pseudo masks for training a segmentation model (seg-model); the seg-model revises the pseudo masks to enhance the reg-model for a better warping in the next iteration. However, there is a key weakness in such dual-model iteration that the spatial misalignment inevitably caused by the reg-model could misguide the seg-model, which makes it converge on an inferior segmentation performance eventually. In this paper, we propose a novel image-aligned style transformation to reinforce the dual-model iterative learning for robust one-shot segmentation of brain tissues. Specifically, we first utilize the reg-model to warp the atlas onto an unlabeled image, and then employ the Fourier-based amplitude exchange with perturbation to transplant the style of the unlabeled image into the aligned atlas. This allows the subsequent seg-model to learn on the aligned and style-transferred copies of the atlas instead of unlabeled images, which naturally guarantees the correct spatial correspondence of an image-mask training pair, without sacrificing the diversity of intensity patterns carried by the unlabeled images. Furthermore, we introduce a feature-aware content consistency in addition to the image-level similarity to constrain the reg-model for a promising initialization, which avoids the collapse of image-aligned style transformation in the first iteration. Experimental results on two public datasets demonstrate 1) a competitive segmentation performance of our method compared to the fully-supervised method, and 2) a superior performance over other state-of-the-art with an increase of average Dice by up to 4.67%. The source code is available at: https://github.com/JinxLv/One-shot-segmentation-via-IST.

Vernacular Architecture: Exploring the Architectural Characteristics and Changes in Traditional Settlements of Rana Tharu from Far Western Nepal

Vernacular architecture and settlements are considered culturally meaningful, socially acceptable, locally harmonized, and ecologically sustainable. The traditional buildings seem to be more connected to the local places, environment, culture, indigenous technology and people. However, with the development of new technologies along with increase in population, replacement of these vernacular buildings and settlements has become a regular phenomenon in many of the nations including Nepal. In this regard, considering the case of vernacular buildings and settlements of Rana Tharu from far western region, this paper aims to (i) examine the characteristics of settlement pattern, house form, construction technology, materials and spaces in Rana Tharu settlements; and (ii) identify the changes and crucial factors behind the acceleration of such changes. The paper applies a qualitative approach through explanatory qualitative analysis for exploring the transformation in architectural style, settlement patterns, house forms, construction techniques, use of spaces, and community livelihoods in the Rana Tharu settlements. Field observations, key informant's interviews and informal discussion with locals were carried out. This research contributes in understanding the factors causing changes in traditional Rana Tharu buildings and settlement, and also highlights on significance of vernacular architecture and settlements.

The impact of technologies on leadership styles in Southeast European countries

This paper focuses on analyzing the range of factors that influence the transformation of leadership styles by technologies. The aim is to evaluate the level of transformation achieved by influence of technologies in certain post-socialist countries in Southeastern Europe (SEE). The theoretical part of the research identifies two sets of factors (endogenous and exogenous) that are important in understanding the influence of technologies on leadership transformation. Key factors were identified, including readiness of organization, perceived-easy of use, customer pressure, law regulations, and leader’s attitude. In the quantitative section of the research, the impact of these factors on the transformation of leadership styles was evaluated. The hypothesis is that these factors significantly impact leadership style transformation in the selected countries, and by understanding their advantages, leaders can improve their performance. The multiple linear regression method was used to research the perception of the impact of selected factors. A mathematical model based on multiple linear regression analysis was created to explain the relationship between the dependent variable (the level of transformation of leadership style) and the selected independent variables. The results have confirmed the validity of the hypothesis and, consequently, the selected factors have the high influence to the transformation of leadership styles in the analyzed countries. The study concludes that the evaluation of the current status of selected factors presents opportunities that leaders should consider to survive the effects of globalization and open markets. The selected factors explain the situation in leadership to a great extent in the observed countries.

Artistic Style Transfer using Deep Learning and Style Fusion- a Review

In recent years, after the study ‘A Neural Algorithm of Artistic Style’ published by Gatys et al. in 2016b, research on style transfer boomed drastically. Style transfer is the process of copying an art style from a ‘style image’ to the contents of the ‘content image’ and producing a ‘draft image’ that is on par with respect to quality expectations. This paper explores different techniques of achieving style transformations namely Style Fusion and Convolutional Neural Networks (CNNs). Although CNNs are the state-of-the-art architecture to tackle cognitive visual tasks, and that they clearly perform much better than most conventional algorithms, the image processing-based style fusion method comes close to the CNN in terms of image output quality and supersedes in terms of time and computation and resources complexity. The procedure of both of these methods has been discussed in detail in this paper and it was concluded that CNNs have a lot more room for improvement that can be facilitated by the availability of better and larger datasets.

SAM-GAN: Supervised Learning-Based Aerial Image-to-Map Translation via Generative Adversarial Networks

Accurate translation of aerial imagery to maps is a direction of great value and challenge in mapping, a method of generating maps that does not require using vector data as traditional mapping methods do. The tremendous progress made in recent years in image translation based on generative adversarial networks has led to rapid progress in aerial image-to-map translation. Still, the generated results could be better regarding quality, accuracy, and visual impact. This paper proposes a supervised model (SAM-GAN) based on generative adversarial networks (GAN) to improve the performance of aerial image-to-map translation. In the model, we introduce a new generator and multi-scale discriminator. The generator is a conditional GAN model that extracts the content and style space from aerial images and maps and learns to generalize the patterns of aerial image-to-map style transformation. We introduce image style loss and topological consistency loss to improve the model’s pixel-level accuracy and topological performance. Furthermore, using the Maps dataset, a comprehensive qualitative and quantitative comparison is made between the SAM-GAN model and previous methods used for aerial image-to-map translation in combination with excellent evaluation metrics. Experiments showed that SAM-GAN outperformed existing methods in both quantitative and qualitative results.

Transforming challenges into opportunities – a case study of Shukan hospital during COVID-19

Learning outcomes This case analysis will make the audience aware of the sudden nature of the crisis, specifically the transboundary crisis. They will be introduced to the concepts of participative leadership and situational leadership. Majorly the audience would learn to remain composed in times of crisis and comprehend how a circumstance can transform a participatory leadership style into a situational leadership style. The case will allow the audience to brainstorm how to take innovative decisions to manage a crisis situation. The primary objectives of the case are as follows: ■ to understand a transboundary crisis and its characteristics to deal with it; ■ to visualise the application of participative and situational leadership and evaluate and analyse the most suitable leadership style during an unknown crisis; ■ to make the students aware of how to remain composed in times of crisis and, for instance, implementing the change of leadership styles per the situation's requirement; and ■ to make students learn to design innovative decision-making strategies to develop a crisis management plan for managing future crises. Case overview/synopsis The case highlights the successful management of the crises faced by a multi-specialty hospital named “Shukan-Caring Lives”, established in 2017 in Vadodara city of Gujarat, India. The case revolves around the Chief Operating Officer, his executive management body and the paramedical and medical staff of the hospital, who dealt with the grave Corona crisis. This case shows how the change in leadership style from participative to situational leadership style could manage the situation by taking brilliant initiatives and providing insightful solutions with a lot of dedication, showing how situational leadership can deal with the transboundary crisis. Complexity academic level The case is an excellent example of a transboundary crisis for any organisation and demonstrates a novel idea of transforming the participative leadership style into a situational leadership style as and when the need arises. The topics of transboundary crisis and management crisis can be taught in strategic management. At the same time, leadership styles and transformation of leadership style can be covered in organisation behaviour. Management development programmes can also include such kinds of cases to give exposure to crisis and its management. The target audience would be PG management students, management executives and senior-level managers in various courses. Supplementary material Teaching notes are available for educators only. Subject code CSS 11: Strategy.

Research on the Transformation of Director Zhang Yimou's Artistic Style

Research on the Transformation of Director Zhang Yimou's Artistic Style

Multi-Modal Multi-Stage Underwater Side-Scan Sonar Target Recognition Based on Synthetic Images

Due to the small sample size of underwater acoustic data and the strong noise interference caused by seabed reverberation, recognizing underwater targets in Side-Scan Sonar (SSS) images is challenging. Using a transfer-learning-based recognition method to train the backbone network on a large optical dataset (ImageNet) and fine-tuning the head network with a small SSS image dataset can improve the classification of sonar images. However, optical and sonar images have different statistical characteristics, directly affecting transfer-learning-based target recognition. In order to improve the accuracy of underwater sonar image classification, a style transformation method between optical and SSS images is proposed in this study. In the proposed method, objects with the SSS style were synthesized through content image feature extraction and image style transfer to reduce the variability of different data sources. A staged optimization strategy using multi-modal data effectively captures the anti-noise features of sonar images, providing a new learning method for transfer learning. The results of the classification experiment showed that the approach is more stable when using synthetic data and other multi-modal datasets, with an overall accuracy of 100%.

Effects of stay-at-home orders on skill requirements in vacancy postings

The COVID-19 pandemic and containment policies have had profound economic impacts on the labor market. Stay-at-home orders (SAHOs) implemented across most of the United States changed the way of people worked. In this paper, we quantify the effect of SAHO durations on skill demands to study how firms adjust labor demand within occupation. We use skill requirement information from the 2018 to 2021 online job vacancy posting data from Burning Glass Technologies, exploit the spatial variations in the SAHO duration, and use instrumental variables to correct for the endogeneity in the policy duration related to local social and economic factors. We find that policy durations have persistent impacts on the labor demand after restrictions are lifted. Longer SAHOs motivate management style transformation from people-oriented to operation-oriented by requiring more of operational and administrative skills and less of personality and people management skills to carry out standard workflows. SAHOs also change the focus of interpersonal skill demands from specific customer services to general communication such as social and writing skills. SAHOs more thoroughly affect occupations with partial work-from-home capacity. The evidence suggests SAHOs change management structure and communication in firms.

The Analysis of Organizational Culture for Improving Corporate Performance at PT. XYZ Discreet

PT. XYZ Discreet is a subsidiary of a State-Owned Enterprise (BUMN) in the manufacturing sector which was founded in 1983. PT. XYZ Discreet contributes to defense & security needs and various industrial products that help in other fields such as transportation and commercial explosives. Organizational culture is essential in determining the success of achieving company goals. Therefore, every company must know what type of culture they are running and what kind of culture is preferred at PT. XYZ Discreet has yet to assess the current and preferred corporate culture. As is well known, BUMN companies and subsidiaries have a new corporate culture, BUMN AKHLAK. Based on an assessment conducted by ACT Consulting, the value of implementing AKHLAK BUMN at PT. XYZ Discreet is not maximized or gets a C value (37.7). The purpose of this study is to provide recommendations for the preferred corporate culture based on survey results. The research methodology used is quantitative data by distributing questionnaires using the Organizational Culture Assessment Instrument (OCAI) to employees of PT. XYZ Discrete. The survey results show that at PT. XYZ Discreet has a current culture Hierarchy culture, but the survey results also state that employees prefer Clan culture as their corporate culture. In other words, they support the culture with kinship, openness, loyalty, mutual trust, and agreement in every activity. From the results of research that has been done at PT. XYZ Discreet, this cultural transformation can be assisted by a bottom-up approach; leaders should have more understanding of employees, the transformation of leadership style, and forming the agent of change for corporate culture. These things aim to improve internal capabilities at PT. XYZ Discreet.