Search Results (194)

Low fruit set in Wurfbainia villosa var. villosa has been a major constraint in its cultivation, with the effective pollination period (EPP) identified as a key factor. In this study, the EPP was assessed for the first time by examining stigma receptivity, style suitability, pollen tube growth rate, and ovule longevity, determined by fluorescence emission microscopy, along with initial fruit set (IFS) determined by sequential hand-pollination of flowers of two cultivars of W. villosa var. villosa under field conditions in Xishuangbanna, South China, in 2022 and 2023. The results showed that the inner surface of the stigma is the receptive region, with receptivity lasting more than three days, as confirmed by pollen adhesion and pollen germination. Style suitability, determined by successful pollen tube entry into the ovule, was maintained for two days; most pollen tubes reached the ovules within one day when stigma receptivity was highest. Ovule longevity persisted for at least three days. The IFS was highest when pollinated at 0–1 days after anthesis (DAA) but dropped sharply to near 0% by 2 DAA. Both EPP estimates, based on its components and IFS, indicated that the EPP is two days, with style suitability being the primary limiting factor. Full article

This research addresses the optimization and deployment of convolutional neural networks for eyeglasses detection on low-power edge devices. Multiple convolutional neural network architectures were trained and evaluated using the FFHQ dataset, which contains annotated eyeglasses in the context of faces with diverse facial features and eyewear styles. Several post-training quantization techniques, including Float16, dynamic range, and full integer quantization, were applied to reduce model size and computational demand while preserving detection accuracy. The impact of model architecture and quantization methods on detection accuracy and inference latency was systematically evaluated. The optimized models were deployed and benchmarked on Raspberry Pi 5 and NVIDIA Jetson Orin Nano platforms. Experimental results show that full integer quantization reduces model size by up to 75% while maintaining competitive detection accuracy. Among the evaluated models, MobileNet architectures achieved the most favorable balance between inference speed and accuracy, demonstrating their suitability for real-time eyeglasses detection in resource-constrained environments. These findings enable efficient on-device eyeglasses detection, supporting applications such as virtual try-ons and IoT-based facial analysis systems. Full article

Background: Data accessibility remains a significant barrier in healthcare AI due to privacy constraints and logistical challenges. Synthetic data, which mimics real patient information while remaining both realistic and non-identifiable, offers a promising solution. Large Language Models (LLMs) create new opportunities to generate high-fidelity clinical conversations between patients and physicians. However, the value of this synthetic data depends on careful evaluation of its realism, accuracy, and practical relevance. Objective: To assess the performance of four leading LLMs: ChatGPT 4.5, ChatGPT 4o, Claude 3.7 Sonnet, and Gemini Pro 2.5 in generating synthetic transcripts of patient–physician interactions in plastic surgery scenarios. Methods: Each model generated transcripts for ten plastic surgery scenarios. Transcripts were independently evaluated by three clinically trained raters using a seven-criterion rubric: Medical Accuracy, Realism, Persona Consistency, Fidelity, Empathy, Relevancy, and Usability. Raters were blinded to the model identity to reduce bias. Each was rated on a 5-point Likert scale, yielding 840 total evaluations. Descriptive statistics were computed, and a two-way repeated measures ANOVA was used to test for differences across models and metrics. In addition, transcripts were analyzed using automated linguistic and content-based metrics. Results: All models achieved strong performance, with mean ratings exceeding 4.5 across all criteria. Gemini 2.5 Pro received mean scores (5.00 ± 0.00) in Medical Accuracy, Realism, Persona Consistency, Relevancy, and Usability. Claude 3.7 Sonnet matched the scores in Persona Consistency and Relevancy and led in Empathy (4.96 ± 0.18). ChatGPT 4.5 also achieved perfect scores in Relevancy, with high scores in Empathy (4.93 ± 0.25) and Usability (4.96 ± 0.18). ChatGPT 4o demonstrated consistently strong but slightly lower performance across most dimensions. ANOVA revealed no statistically significant differences across models (F(3, 6) = 0.85, p = 0.52). Automated analysis showed substantial variation in transcript length, style, and content richness: Gemini 2.5 Pro generated the longest and most emotionally expressive dialogues, while ChatGPT 4o produced the shortest and most concise outputs. Conclusions: Leading LLMs can generate medically accurate, emotionally appropriate synthetic dialogues suitable for educational and research use. Despite high performance, demographic homogeneity in generated patients highlights the need for improved diversity and bias mitigation in model outputs. These findings support the cautious, context-aware integration of LLM-generated dialogues into medical training, simulation, and research. Full article

Precise and real-time estimation of the robotic arm’s position on the patient’s side is essential for the success of remote robotic surgery in Tactile Internet (TI) environments. This paper presents a prediction model based on the Transformer-based Informer framework for accurate and efficient position estimation, combined with a Four-State Hidden Markov Model (4-State HMM) to simulate realistic packet loss scenarios. The proposed approach addresses challenges such as network delays, jitter, and packet loss to ensure reliable and precise operation in remote surgical applications. The method integrates the optimization problem into the Informer model by embedding constraints such as energy efficiency, smoothness, and robustness into its training process using a differentiable optimization layer. The Informer framework uses features such as ProbSparse attention, attention distilling, and a generative-style decoder to focus on position-critical features while maintaining a low computational complexity of $\mathcal{O}(LlogL)$. The method is evaluated using the JIGSAWS dataset, achieving a prediction accuracy of over 90% under various network scenarios. A comparison with models such as TCN, RNN, and LSTM demonstrates the Informer framework’s superior performance in handling position prediction and meeting real-time requirements, making it suitable for Tactile Internet-enabled robotic surgery. Full article

Hybrid evolutionary approaches have gained significant attention for solving complex optimization problems, but their potential for optimizing the low-dimensional latent space of generative adversarial networks (GANs) remains underexplored. This paper proposes a novel improved crisscross optimization (ICSO) algorithm, a hybrid evolutionary approach that integrates crisscross optimization and perturbation mechanisms to find the suitable latent vector. The ICSO algorithm treats the quality and diversity as separate objectives, balancing them through a normalization strategy, while a gradient regularization term (i.e., GP) is introduced into the discriminator’s objective function to stabilize training and mitigate gradient-related issues. By combining the global and local search capabilities of particle swarm optimization (PSO) with the rapid convergence of crisscross optimization, ICSO efficiently explores and exploits the latent space. The extensive experiments demonstrate that ICSO outperforms state-of-the-art algorithms in optimizing the latent space of various classical GANs across multiple datasets. Furthermore, the practical applicability of ICSO is validated through its integration with StyleGAN3 for generating unmanned aerial vehicle (UAV) images, showcasing its effectiveness in real-world engineering applications. This work not only advances the field of GAN optimization but also provides a robust framework for applying hybrid evolutionary algorithms to complex generative modeling tasks. Full article

The evolution of tall buildings has been shaped by distinct architectural styles, beginning around 1875 and progressing through various stylistic architectural movements. These changes were driven by advancements in structural engineering and digital design technologies, leading to greater experimentation with form and function. Energy and resource conservation of the late 20th century instigated a noteworthy focus on sustainability. Beyond that, the early 21st century saw a significant shift toward a new breed of tall buildings, a suitable architectural vocabulary for “high-performance” tall buildings, in which sustainability with a focus on energy efficiency is joined with the performance of other active and passive functional systems. This paper presents an overview of high-performance tall buildings by exploring key technologies, materials, innovations, safety, durability, and indoor environmental quality. Strategies that have emerged to address skyscrapers’ environmental and economic challenges are also crucial in such a building. It highlights the importance of optimizing and integrating building systems, improving energy efficiency, minimizing resource consumption, and ensuring long-term occupant health and productivity. Furthermore, this study identifies five key dimensions—structural materials and systems, energy-efficient design, high-performance façades, performance monitoring, and integrating building services systems—demonstrating how these factors contribute to environment-conscious urban development and resilient architectural and engineering design. It is concluded that these buildings are poised to redefine urban environments by leveraging advanced technologies, AI-driven management, IoT interconnectivity, health-focused elements, and climate resilience. Also, tall, high-performance buildings will be increasingly automated to an unknown limit, and AI will play a prominent role in the future. Full article

Apple production in Western Norway faces challenges due to climatic constraints and varying phenology. It is essential for cultivars to adapt to regional ecological factors, while suitable pollinators are necessary for successful cultivation. This study examined the reproductive biology of two newly introduced apple cultivars, ‘Eden’ (Wursixo) and ‘Fryd’ (Wuranda), over two years (2022–2023). Key qualitative and quantitative parameters of reproductive biology were analyzed, including in vitro pollen germination, pollen tube growth within the style and ovary locules, flowering overlap time, and fruit set. The study involved cross-pollination between the pollen recipient cultivars ‘Eden’ and ‘Fryd’, with various pollenizers: ‘Rubinstep’, ‘Red Aroma’, ‘Elstar’, ‘Asfari’ and ‘Professor Sprenger’, as well as self-pollination and open pollination. According to the results from the progamic phase of fertilization and fruit set, the cultivars ‘Rubinstep’, ‘Asfari’, and ‘Fryd’ were the best pollenizers for ‘Eden’. In contrast, ‘Rubinstep’, ‘Eden’, and ‘Elstar’ were the best pollenizers for ‘Fryd’. Looking only at the overlapping of the flowering time between pollen recipient and pollen donor, ‘Professor Sprenger’ and ‘Fryd’ were the best pollenizers for ‘Eden’, while ‘Professor Sprenger’ and ‘Eden’ were good pollenizers for ‘Fryd’. Full article

In recent years, virtual reality (VR) has emerged as a pivotal technology in the field of education, offering immersive and interactive experiences that have the potential to significantly enhance teaching and learning processes. However, for educational designers, teachers, and lecturers who lack advanced information technology skills, identifying and evaluating suitable VR applications can be a complex and labour-intensive endeavour. This article provides a non-technical, action-oriented overview of commercially available VR solutions designed for social skills training in education. To identify relevant providers, we conducted a four-phase snowball search from April to October 2024, combining academic literature reviews, expert interviews, web searches, and demo testing. This process led to the identification of ten VR providers suitable for social and communication skills development. The selected applications were analyzed across ten key criteria, including communication style, avatar design and customization, technological requirements, language support, and the availability of content libraries and authoring tools. The results reveal significant variation among providers in terms of pedagogical flexibility, technical accessibility, and degrees of user customization. By mapping these differences, the article offers a practical starting point for educational stakeholders seeking VR solutions that align with their instructional goals. Rather than offering an exhaustive review, the study presents a structured orientation to facilitate informed decision-making in selecting and implementing VR tools for social skills training. Full article

An analysis of the practice path and methodology system of Artificial Intelligence Generated Content (AIGC) technology has been conducted in the field of inheritance and innovation of boundary paintings from Song Dynasty. This paper aims to provide valuable reference and guidance for the application of AI technology in Song Dynasty boundary painting (Song painting) in the interior decoration design, so as to promote the effective integration of traditional aesthetics and modern design concepts. Firstly, the natural processing language model is used to generate the index layer suitable for the indoor soft decoration style of Song paintings, and the Analytic Hierarchy Process weight classification is used to select the cue words of the generated image. Secondly, Midjourney is used to generate Song Dynasty style images for keywords. Finally, Stable Diffusion control model is used to transfer the style of Song painting elements to interior decoration design. AIGC technology can effectively generate images with the style of Song painting elements, and play a unique role in style transfer and pattern design. It provides an innovative path for the integration of traditional art and modern design, and provides a wealth of possibilities for the modern application of Song painting. AIGC technology has significant potential in the inheritance and innovation of Song painting, which can bring new ideas and methods for interior decoration design, and contribute to the wide application and innovative development of Song painting art in the field of modern design. Full article

In the Programme for International Student Assessment (PISA) 2015, students from four cities/provinces in mainland China performed worse in collaborative problem solving (CPS) than in other subjects. While student collaboration has been widely implemented in Chinese classrooms for over two decades, empirical research on teachers’ roles and interventions remains quite scarce. Influenced by international educational reform in the 21st century, educators have developed and made widespread use of open-ended tasks, perceived as more suitable for CPS, during mathematics lessons. In this study, we investigate the effect of teacher intervention during pair and small group CPS using a quasi-experiment with four teachers from eight classes. We then selected typical cases and analysed their effect on task performance regarding intervention focus and means. The result showed that three of the four teachers’ interventions proved effective. The most and least effective teachers were selected for the case study. We discuss teacher intervention’s effect in emphasising social activities and diagnosing. Considering the difference in authority between teachers in Chinese/Western classrooms, we discuss intervention styles and offer suggestions for choosing and carefully implementing appropriate forms of collaborative activities. Full article

In addition to the widely recognized benefits of reducing carbon emissions and protecting the environment, the authors believe that bus electrification has potential advantages in enhancing driving safety, improving passenger comfort, and reducing driver fatigue—areas that have not yet been sufficiently studied and emphasized. Safety and comfort are fundamental objectives in the continuous development of transportation systems. They are directly and closely related to both passengers and drivers and are among the top priorities when individuals choose their mode of transportation. Therefore, these aspects deserve broader and more in-depth attention and research. This study aims to identify the potential advantages of route bus electrification in terms of safety and comfort. The results of a passive experiment on the speed profile of buses operating on actual routes are presented here. Firstly, we focus on the acceleration/deceleration at the starting/stopping stops, specifically for regular-route buses, and obtain the following information: I. Starting acceleration from a bus stop is particularly strong in the second half of the acceleration process, being suitable for motor-driven vehicles. II. The features of the stopping deceleration at a bus stop are “high intensity” and “low dispersion”, with the latter enabling the refinement of regenerative settings and significantly lowering electricity economy during electrification. And we compare the speed profile of an electric bus with those of a diesel bus and obtain the following information: III. Motor-driven vehicles offer the advantages of “high acceleration performance” and “no gear shifting”, making them particularly suitable for the high-intensity acceleration required when route buses depart from stations. This not only simplifies driving operations but also enhances lane-changing safety. And by calculating and analyzing the jerk amount, we could quantitatively demonstrate the comfortable driving experience while riding on this type of bus where there is no shock due to gear shifting. IV. While the “high acceleration performance” of motor-driven vehicles produces “individual differences in the speed change patterns”, this does not translate to “individual differences in electricity consumption”, owing to the characteristics of this type of vehicle. With engine-driven vehicles, measures such as “slow acceleration” and “shift up early” are strongly encouraged to realize eco-driving, and any driving style that deviates from these measures is avoided. However, with motor-driven vehicles, the driver does not need to be too concerned about the speed change patterns during acceleration. This characteristic also suggests a benefit in terms of the electrification of buses. Full article

Sign language plays a crucial role in bridging communication barriers within the Deaf community. Recognizing sign language sentences remains a significant challenge due to their complex structure, variations in signing styles, and temporal dynamics. This study introduces an innovative sign language sentence recognition (SLSR) approach using Hybrid Graph Embedding and Adaptive Convolutional Networks (HGE-ACN) specifically developed for single-handed wearable glove devices. The system relies on sensor data from a glove with six-axis inertial sensors and five-finger curvature sensors. The proposed HGE-ACN framework integrates graph-based embeddings to capture dynamic spatial–temporal relationships in motion and curvature data. At the same time, the Adaptive Convolutional Networks extract robust glove-based features to handle variations in signing speed, transitions between gestures, and individual signer styles. The lightweight design enables real-time processing and enhances recognition accuracy, making it suitable for practical use. Extensive experiments demonstrate that HGE-ACN achieves superior accuracy and computational efficiency compared to existing glove-based recognition methods. The system maintains robustness under various conditions, including inconsistent signing speeds and environmental noise. This work has promising applications in real-time assistive tools, educational technologies, and human–computer interaction systems, facilitating more inclusive and accessible communication platforms for the deaf and hard-of-hearing communities. Future work will explore multi-lingual sign language recognition and real-world deployment across diverse environments. Full article

Bamboos, as imperative vegetations in Chinese traditional gardens, also significantly influenced the recently originated Neo-Chinese-style landscape in China, and their habitat ranges have been profoundly impacted by global climate warming. Current studies on the distribution dynamics of bamboo reveal existent gaps in assessing the suitable distribution area of Neo-Chinese-style landscapes. In this study, we calculated the habitat ranges of two widely distributed bamboo genera (Phyllostachys and Bambusa) based on the optimal MaxEnt model, predicted their future (2050s, 2070s and 2090s) distributions under different climate scenarios (SSP1-2.6 and SSP5-8.5), and assessed the suitable distribution area of the Neo-Chinese-style landscape according to the distribution union of two bamboo genera. The results showed that the optimal MaxEnt model exhibited high evaluation indices (AUC > 0.90) for the two bamboo genera. The habitat ranges of bamboo genera were significantly influenced by the minimum temperature of the coldest month and would expand northwardly in the future. The suitable distribution area of Neo-Chinese-style landscapes covered about 71.3% cities of China, which would expand 5.9%–8.7% of cities and 10%–18.7% of cities under the SSP1-2.6 climate scenario and the SSP5-8.5 climate scenario, respectively. The suitable distributions are mainly located in the southeast part of China. This study advanced our understanding of the restriction of bamboo to the distribution of the Neo-Chinese-style landscape and provided valuable insights and a scientific basis for landscape construction in different areas of China. Full article

Onion (Allium cepa L.) plants produce umbellate inflorescences that bear dozens of hermaphrodite flowers. Outcrossing is encouraged by a strong protandrous dichogamy. Honeybees are reported as the most suitable pollinator for this crop due to their efficiency in pollen transfer. However, bee activity does not always guarantee pollination success and an adequate seed set. This work aimed to analyze the developmental changes of the pistil along the flower lifespan to determine the reasons for common failures in the onion seed set. Studies were carried out in two productive seasons, in a male sterile line and a fertile line, using two different pollination procedures. The results showed that the stigma papillae did not fully develop until 4 days after anthesis (DAA), reducing the stigma receptivity to only 4 days. The style did not reach its maximum length until 8 DAA, when ovule fertility was achieved. Eight days were needed for the pollen tube to grow, reach, and fertilize the ovules. The results were consistent over two consecutive years. The short period of stigmatic receptivity and strong protandrous dichogamy limit the duration of the effective pollination period and could explain the poor cropping performance in obtaining seeds from some onion hybrid lines. Full article

We develop a deep reinforcement learning (RL) framework for an optimal market-making (MM) trading problem, specifically focusing on price processes with semi-Markov and Hawkes Jump-Diffusion dynamics. We begin by discussing the basics of RL and the deep RL framework used; we deployed the state-of-the-art Soft Actor–Critic (SAC) algorithm for the deep learning part. The SAC algorithm is an off-policy entropy maximization algorithm more suitable for tackling complex, high-dimensional problems with continuous state and action spaces, like those in optimal market-making (MM). We introduce the optimal MM problem considered, where we detail all the deterministic and stochastic processes that go into setting up an environment to simulate this strategy. Here, we also provide an in-depth overview of the jump-diffusion pricing dynamics used and our method for dealing with adverse selection within the limit order book, and we highlight the working parts of our optimization problem. Next, we discuss the training and testing results, where we provide visuals of how important deterministic and stochastic processes such as the bid/ask prices, trade executions, inventory, and the reward function evolved. Our study includes an analysis of simulated and real data. We include a discussion on the limitations of these results, which are important points for most diffusion style models in this setting. Full article