Search Results (173)

This study investigates how Greek professionals formulate upward requests and simultaneously manage rapport and workplace identity within hierarchical exchanges. The data comprise 400 written requests elicited through a discourse–completion task from 100 participants, supplemented by follow-up interviews. Integrating pragmatic perspectives on request mitigation with Spencer-Oatey’s Rapport-Management model and a social constructionist perspective on identity, the analysis reveals a distinctive “direct-yet-mitigated” style: syntactically direct head acts (typically want- or need-statements) various mitigating devices. This mitigation enables speakers to preserve superiors’ face, assert entitlement, and invoke shared corporate goals in a single move. Crucially, rapport work is intertwined with identity construction. Strategic oscillation between deference and entitlement projects four recurrent professional personae: the deferential subordinate, the competent and deserving employee, the cooperative team-player, and the rights-aware negotiator. Speakers shift among these personae to calibrate relational distance, demonstrating that rapport management functions not merely as a politeness calculus but as a resource for dynamic identity performance. This study thus bridges micro-pragmatic choices and macro social meanings, showing how linguistic mitigation safeguards interpersonal harmony while scripting desirable workplace selves. Full article

Traditional Foreign Object Debris (FOD) detection methods face challenges such as difficulties in large-size data acquisition and the ineffective application of detection algorithms with high accuracy. In this paper, image data augmentation was performed using generative adversarial networks and diffusion models, generating images of monitoring areas under different environmental conditions and FOD images of varied types. Additionally, a three-stage image blending method considering size transformation, a seamless process, and style transfer was proposed. The image quality of different blending methods was quantitatively evaluated using metrics such as structural similarity index and peak signal-to-noise ratio, as well as Depthanything. Finally, object detection models with a similarity distance strategy (SimD), including Faster R-CNN, YOLOv8, and YOLOv11, were tested on the dataset. The experimental results demonstrated that realistic FOD data were effectively generated. The Structural Similarity Index Measure (SSIM) and Peak Signal-to-Noise Ratio (PSNR) of the synthesized image by the proposed three-stage image blending method outperformed the other methods, reaching 0.99 and 45 dB. YOLOv11 with SimD trained on the augmented dataset achieved the mAP of 86.95%. Based on the results, it could be concluded that both data augmentation and SimD significantly improved the accuracy of FOD detection. Full article

Automotive intelligent agents are increasingly collecting facial data for applications such as driver behavior monitoring and identity verification. These excessive collections of facial data bring serious risks of sensitive information leakage to autonomous driving. Facial information has been explicitly required to be anonymized, but the availability of most desensitized facial data is poor, which will greatly affect its application in autonomous driving. This paper proposes an automotive sensitive information anonymization method with high-quality generated facial images by considering the data availability under privacy protection. By comparing K-Same and Generative Adversarial Networks (GANs), this paper proposes a hierarchical self-attention mechanism in StyleGAN3 to enhance the feature perception of face images. The synchronous regularization of sample data is applied to optimize the loss function of the discriminator of StyleGAN3, thereby improving the convergence stability of the model. The experimental results demonstrate that the proposed facial desensitization model reduces the Frechet inception distance (FID) and structural similarity index measure (SSIM) by 95.8% and 24.3%, respectively. The image quality and privacy desensitization of the facial data generated by the StyleGAN3 model have been fully verified in this work. This research provides an efficient and robust facial privacy protection solution for autonomous driving, which is conducive to promoting the security guarantee of automotive data. Full article

Hyperspectral image cross-scene classification (HSICC) tasks are confronted with tremendous challenges due to spectral shift phenomena across scenes and the tough work of obtaining labels. Unsupervised domain adaptation has proven its effectiveness in tackling this issue, but it has a fatal limitation of intending to narrow the disparity between source and target domains by utilizing fully labeled source data and unlabeled target data. However, it is costly even to attain labels from source domains in many cases, rendering sufficient labeling as used in prior work impractical. In this work, we investigate an extreme and realistic scenario where unsupervised domain adaptation methods encounter sparsely labeled source data when handling HSICC tasks, namely, few-shot unsupervised domain adaptation. We propose an end-to-end refined bi-directional prototypical contrastive learning (RBPCL) framework for overcoming the HSICC problem with only a few labeled samples in the source domain. RBPCL captures category-level semantic features of hyperspectral data and performs feature alignment through in-domain refined prototypical self-supervised learning and bi-directional cross-domain prototypical contrastive learning, respectively. Furthermore, our framework introduces the class-balanced multicentric dynamic prototype strategy to generate more robust and representative prototypes. To facilitate prototype contrastive learning, we employ a Siamese-style distance metric loss function to aggregate intra-class features while increasing the discrepancy of inter-class features. Finally, extensive experiments and ablation analysis implemented on two public cross-scene data pairs and three pairs of self-collected ultralow-altitude hyperspectral datasets under different illumination conditions verify the effectiveness of our method, which will further enhance the practicality of hyperspectral intelligent sensing technology. Full article

The stylistic emulation of historical building facades presents significant challenges for artificial intelligence (AI), particularly for complex and data-scarce styles like Krakow’s Eclecticism. This study aims to develop a methodological framework for a “typological transcoding” of style that moves beyond mere visual mimicry, which is crucial for heritage preservation and urban renewal. The proposed methodology integrates architectural typology with Low-Rank Adaptation (LoRA) for fine-tuning a Stable Diffusion (SD) model. This process involves a typology-guided preparation of a curated dataset (150 images) and precise control of training parameters. The resulting typologically guided LoRA-tuned model demonstrates significant performance improvements over baseline models. Quantitative analysis shows a 24.6% improvement in Fréchet Inception Distance (FID) and a 7.0% improvement in Learned Perceptual Image Patch Similarity (LPIPS). Furthermore, qualitative evaluations by 68 experts confirm superior realism and stylistic accuracy. The findings indicate that this synergy enables data-efficient, typology-grounded stylistic emulation, highlighting AI’s potential as a creative partner for nuanced reinterpretation. However, achieving deeper semantic understanding and robust 3D inference remains an ongoing challenge. Full article

Seoul systematically removed all graveyards that once lay within the city and its surrounding areas, a phenomenon notably distinct from urban development patterns in other parts of the world. After the Korean War, refugees and migrants poured into the devastated capital. In this postwar environment, cemeteries—traditionally sites of mourning and death—transformed into spaces of survival for displaced populations. With the military demarcation line preventing their return home, refugees began to envision their lost hometowns as “absent places”: unattainable utopias, idealized lands where all beauty resides—the very origin and endpoint of life. In contrast, Seoul, where they were forced to settle, became a “dystopia,” stripped of sanctity. Over time, however, the next generation reinterpreted this dystopia, gradually transforming it into a heterotopia. As Seoul’s urban landscape expanded, this heterotopia evolved into a Christian paradise. The second generation, having never experienced the trauma of displacement, found the newly constructed city comfortable and secure. Reinforced concrete buildings and asphalt roads became symbolic of paradise. The development of Gangnam—famously captured in Psy’s global hit “Gangnam Style”—represents a belated cultural revolution among younger generations in modern South Korea and exemplifies the transformation into a concrete paradise. Full article

Accurate glaucoma diagnosis relies on precise segmentation of the optic disc (OD) and optic cup (OC) in retinal images. However, despite the development of numerous automatic segmentation models, the lack of annotations in the target domain and domain shift among datasets continue to limit their segmentation performance. To address these issues, we propose a Causal Self-Supervised Network (CSSN) that leverages self-supervised learning to enhance model performance. First, we construct a Structural Causal Model (SCM) and employ backdoor adjustment to convert the conventional conditional distribution into an interventional distribution, effectively severing the influence of style information on feature extraction and pseudo-label generation. Subsequently, the low-frequency components of source and target domain images are exchanged via Fourier transform to simulate cross-domain style transfer. The original target images and their style-transferred counterparts are then processed by a dual-path segmentation network to extract their respective features, and a confidence-based pseudo-label fusion strategy is employed to generate more reliable pseudo-labels for self-supervised learning. In addition, we employ adversarial training and cross-domain contrastive learning to further reduce style discrepancies between domains. The former aligns feature distributions across domains using a feature discriminator, effectively mitigating the adverse effects of style inconsistency, while the latter minimizes the feature distance between original and style-transferred images, thereby ensuring structural consistency. Experimental results demonstrate that our method achieves more accurate OD and OC segmentation in the target domain during testing, thereby confirming its efficacy in cross-domain adaptation tasks. Full article

With the advancement of agricultural automation, the demand for road recognition and understanding in agricultural machinery autonomous driving systems has significantly increased. To address the scarcity of instance segmentation data for rural roads and rural unstructured scenes, particularly the lack of support for high-resolution and fine-grained classification, a 20-class instance segmentation dataset was constructed, comprising 10,062 independently annotated instances. An improved StyleGAN2-ADA data augmentation method was proposed to generate higher-quality image data. This method incorporates a decoupled mapping network (DMN) to reduce the coupling degree of latent codes in W-space and integrates the advantages of convolutional networks and transformers by designing a convolutional coupling transfer block (CCTB). The core cross-shaped window self-attention mechanism in the CCTB enhances the network’s ability to capture complex contextual information and spatial layouts. Ablation experiments comparing the improved and original StyleGAN2-ADA networks demonstrate significant improvements, with the inception score (IS) increasing from 42.38 to 77.31 and the Fréchet inception distance (FID) decreasing from 25.09 to 12.42, indicating a notable enhancement in data generation quality and authenticity. In order to verify the effect of data enhancement on the model performance, the algorithms Mask R-CNN, SOLOv2, YOLOv8n, and OneFormer were tested to compare the performance difference between the original dataset and the enhanced dataset, which further confirms the effectiveness of the improved module. Full article

As global initiatives to reduce greenhouse gas emissions and combat climate change expand, electric vehicles (EVs) powered by fuel cells and lithium-ion batteries are gaining global recognition as solutions for sustainable transportation due to their high energy conversion efficiency. Considering the driving range limitations of battery electric vehicles (BEVs) and the low efficiency of internal combustion engines (ICEs), fuel cell hybrid vehicles offer a compelling alternative for long-distance, low-emission driving with less refuelling time. To facilitate their wider scale adoption, it is essential to understand their energy performance through models that consider external weather effects, driving styles, road gradients, and their simultaneous interaction. This paper presents a microlevel, multicriteria assessment framework to investigate the performance of BEVs, fuel cell electric vehicles (FCEVs), and hybrid electric vehicles (HEVs), with a focus on energy consumption, drive systems, and emissions. Simulation models were developed using MATLAB 2021a Simulink environment, thus enabling the integration of standardized driving cycles with real-world wind and terrain variations. The results are presented for various trip scenarios, employing quantitative and qualitative analysis methods to identify the most efficient vehicle configuration, also validated through the simulation of three commercial EVs. Predictive modelling approaches are utilized to estimate a vehicle’s performance under unexplored conditions. Results indicate that trip conditions have a significant impact on the performance of all three vehicles, with HEVs emerging as the most efficient and balanced option, followed by FCEVs, making them strong candidates compared with BEVs for broader adoption in the transition toward sustainable transportation. Full article

Background: Limited data exist on managing type 1 diabetes mellitus (T1DM) during long-distance endurance events such as marathons. This study aimed to assess glycemic control and participant safety during a marathon. Methods: Five men with T1DM, participating in the 22nd Poznan Marathon, were recruited. They completed health questionnaires and received training on glycemic management. Their physical capacity was assessed (including maximal oxygen uptake on a cycle ergometer). Participants reduced their insulin doses and consumed breakfast 2.5–3 h before the race. During the marathon, self-monitoring blood glucose (SMBG) and ketone levels were measured at five checkpoints (start, 10 km, 19 km, 30 km, and finish). The medical team followed a pre-approved protocol, providing carbohydrate and fluid supplementation as needed. Glycemia was monitored by two continuous glucose monitoring (CGM) systems (FreeStyle Libre 2 and Dexcom G6) and SMBG. Results: The participants’ median age was 44 years (34–48), with a diabetes duration of 10 years (6–14), and a BMI of 22.5 kg/m² (22.0–23.3). All finished the marathon in an average time of 4:02:56 (±00:43:11). Mean SMBG was 125.6 (±43.5) mg/dL, while CGM readings were 149.6 (±17.9) mg/dL (FreeStyle Libre 2) and 155.4 (±12.9) mg/dL (Dexcom G6). One participant experienced prolonged hypoglycemia undetected by CGM, whereas another developed symptomatic hypoglycemia between SMBG measurements. Conclusions: Safe marathon completion in people with T1DM requires individualized insulin dose adjustments, appropriate carbohydrate supplementation, and dedicated medical support at checkpoints. Combining CGM with periodic SMBG measurements further enhances safety and helps to detect potential glycemic excursions. Full article

With the rapid expansion of the aviation industry, pier-style departure lounges have become increasingly prevalent in modern airport terminals. Unlike traditional long enclosures—such as corridors, tunnels, and subway stations—airport terminal piers feature unique geometries, volumes, and interior finishes which complicate sound propagation. To address the paucity of objective acoustic data in these expansive environments, this study performed in situ measurements of impulse responses and sound pressure levels in two piers with distinct shapes and volumes within the same terminal. Key acoustic parameters, including the A-weighted equivalent continuous sound pressure level (L_Aeq), early decay time (EDT), reverberation time (T₃₀), definition (D₅₀), and speech transmission index (STI), were analyzed. The results reveal that EDT and T₃₀ increase significantly with distance from the sound source, while D₅₀ and STI decrease correspondingly. Specifically, compared to Pier B, which has a smaller cross-sectional area and a single-sided layout, Pier A, characterized by a larger cross-sectional area and a double-sided layout, exhibits a faster sound attenuation when the receiver is positioned closer to the source and a longer reverberation time when the receiver is farther from the source. Notably, STI does not differ significantly between the two piers. These findings enhance the understanding of acoustic behavior in large-span, elongated airport piers and provide valuable guidance for optimizing the acoustic environment of departure lounges to improve passenger comfort. Full article

In an era of rapid market shifts and technological disruption, the success of organizational change rests on the ability of leaders to navigate complex cultural dynamics. This study explores how culturally adaptive leadership can drive employee outcomes in Sierra Leone’s commercial banking sector during periods of change. By integrating transformational and transactional leadership styles with Hofstede’s cultural dimensions theory, which focuses on power distance (respect for authority) and uncertainty avoidance (preference for structure), this research examines how these cultural values influence the relationship between leadership approaches and employee outcomes. Using a cross-sectional design, data were collected from 820 employees across commercial banks in Sierra Leone, with data analyzed using structural equation modeling (SEM). The findings reveal that transformational leadership significantly enhances employee outcomes, specifically in high power distance environments where authority is deeply respected, while transactional leadership proves more effective in high uncertainty avoidance settings, where clear structure and predictability are paramount. The study highlights the complementary nature of these leadership styles, suggesting that effective leaders must adapt their strategies to the cultural context to drive performance. While the cross-sectional design limits causal inference, this research underscores the critical importance of culturally adaptive leadership, recognizing how cultural dimensions shape behavior and promote sustained success during change. Full article

This study investigates the relationship between sustainable leadership styles and conflict management strategies within the context of Greek Public Sector. Specifically, it examines how collaborative, transformational, and authoritarian leadership styles impact workplace conflict resolution. The research adopts a case study methodology, focusing on Departments of Public Works in Greece, where data were collected through questionnaires. The analysis involved quantitative methods, including exploratory factor analysis (EFA), to examine the relationship between leadership styles and conflict management techniques. Results indicate that collaborative leadership is strongly associated with higher employee satisfaction and more effective conflict resolution, particularly in organizations with flat hierarchical structures. Transformational leadership fosters trust and open communication, which further enhance conflict resolution. On the other hand, authoritarian leadership styles correlate with increased workplace tension, lower satisfaction, and less effective conflict management, especially in high power-distance environments. The study also highlights cultural factors, such as the Greek emphasis on interpersonal relationships, as critical influences on leadership effectiveness. These findings underline the need for culturally adaptive and sustainable leadership strategies and provide practical recommendations for promoting harmony and productivity in Greek organizations. Full article

In recent years, an increasing number of accidents at ski resorts have raised significant safety concerns. To address these issues, it is essential to understand skiing traffic and the underlying dynamics. We collected 225 trajectories, which were analyzed after a correction process. To obtain a quantitative classification of typical trajectories we focus on three main quantities: turning points, minimum distance, and the fundamental diagram. Our objective was to analyze these trajectories in depth and identify key statistical properties. Our findings indicate that three factors—turning angle, curvature, and velocity change—can be used to accurately identify turning points and classify skiers’ movement styles. We found that aggressive skiers tend to exhibit larger and less stable turning angles, while conservative skiers demonstrate a more controlled style, characterized by smaller, more stable turns. This is consistent with observations made for the distribution of the minimum distance to other skiers. Furthermore, we have derived a fundamental diagram which is an important characteristic of any traffic system. It is found share more similarities with the fundamental diagram of ant trails than those of highway traffic. Full article

Previously, it was suggested that the “persona-driven” approach can contribute to producing sufficiently diverse synthetic training data for Large Language Models (LLMs) that are currently about to run out of real natural language texts. In our paper, we explore whether personas evoked from LLMs through HCI-style descriptions could indeed imitate human-like differences in authorship. For this end, we ran an associative experiment with 50 human participants and four artificial personas evoked from the popular LLM-based services: GPT-4(o) and YandexGPT Pro. For each of the five stimuli words selected from university websites’ homepages, we asked both groups of subjects to come up with 10 short associations (in Russian). We then used cosine similarity and Mahalanobis distance to measure the distance between the association lists produced by different humans and personas. While the difference in the similarity was significant for different human associators and different gender and age groups, neither was the case for the different personas evoked from ChatGPT or YandexGPT. Our findings suggest that the LLM-based services so far fall short at imitating the associative thesauri of different authors. The outcome of our study might be of interest to computer linguists, as well as AI/ML scientists and prompt engineers. Full article