This study tackles the challenge of optimizing aesthetics, energy use, and material sustainability for eco-friendly visual identity systems in dynamically operated green buildings. A “Metric-Simulation-Control-Evaluation” closed-loop framework was developed. It integrates multi-objective entropy weight assessment, light-thermal simulation, and a vector quantized-variational autoencoder-based color-footprint material library, supporting the generation and selection of sustainable visual solutions. In operation, an adaptive dimming system using the twin delayed deep deterministic policy gradient reinforcement learning algorithm enabled real-time coordinated control of lighting and shading. Experiments demonstrated annual electricity savings of 12,412 kWh and increased user satisfaction to 4.5 (on a five-point scale), confirming synergistic gains in energy conservation, environmental impact, and experience. This work provides a practical pathway for dynamic, intelligent ecological visual interaction in green buildings.
Immersive extended reality environments increasingly require adaptive simulations responsive to users’ affective and behavioral states. Existing systems often rely on rule-based interactions, limiting presence and flow. In this study, the author presents an artificial intelligence-driven ambient intelligence framework for context-aware extended reality that combines affective semantic–action mapping, reinforcement learning-based multisensory control, flow prediction, and adaptive narrative generation in a closed-loop architecture. A 60-participant experiment using heart rate variability, galvanic skin response, interaction logs, and experience measures showed significant improvements in presence and flow, with reduced cognitive workload and stress. Results demonstrated the promise of context-aware AI for enhancing immersive human–computer interaction.
This article presents a virtual reality landscape sculpture framework that applied ambient intelligence to transform conventional static landscape sculptures into context-aware interactive installations. Based on multi-source environmental sensors and semantic annotation, the system supported real-time three-way interaction between environment, sculptures, and users. It integrated the multisensory fusion rendering algorithm for synchronized visual–auditory–tactile feedback and the gesture eye synergy model for implicit natural interaction, with 42 ms response latency. Tests with 48 participants showed that the virtual reality–landscape sculpture framework achieved significantly higher immersion depth (0.82 versus 0.66), emotion valence (0.79 versus 0.63), and knowledge retention (0.68 versus 0.55) than mobile augmented reality, all at p < 0.01. The 30-day knowledge retention rate reached 42%, 2.3 times that of traditional static viewing. This study offers a reproducible technical approach for immersive public art design and sustainable digital conservation of cultural heritage.
This study investigates the design and application of a remote monitoring system for green logistics transportation devices based on image control and processing modules. By integrating GPS positioning, wireless communication, and image processing technologies, the system enables real-time monitoring of vehicle temperature, humidity, and location data, as well as anomaly alerts and remote scheduling. The study develops a framework for a green logistics transportation system that optimizes resource allocation and transport routes to reduce operational costs. Experimental results demonstrate that the system performs effectively in environmental monitoring and node conflict management, significantly enhancing the automation and environmental sustainability of logistics transportation. Furthermore, the integration of dynamic frame slot and fixed frame delay algorithms effectively reduces node conflicts and improves system stability. The findings provide a practical technological solution for promoting the intelligent and sustainable development of green logistics.
Smart city big data governance is encumbered by data siloing, poor quality, and inefficient sharing, restricting ambient computing integration. This study applies artificial intelligence technologies—federated learning, knowledge graphs, and deep learning—to data cleaning, cross-departmental model sharing, and decision making. It proposes a technology–institution–context integrated model. Empirically analyzing three Chinese pilot cities from 2020 to 2023 shows that the AI strategy significantly reduced data cleaning time, elevated anomaly detection accuracy to over 95%, improved data standardization from approximately 30% to over 80% (with City C reaching 90.5%), increased cross-departmental shared data by approximately 5.5 to over 6 times, cut governance costs by approximately 62%, and lowered decision errors to approximately 1/3 of that of traditional methods. Challenges persist: inconsistent cross-regional standards, a lack of unified AI norms, and weak algorithm interpretability. This study enriches ambient computing governance theory and provides a practical framework for urban managers.
This paper proposes a deep learning-based approach for bridge health monitoring, addressing the inefficiencies and limitations of conventional inspection methods. As critical transportation infrastructure, bridge conditions directly impact public safety. Traditional monitoring techniques, however, are often labor-intensive, time-consuming, and costly, failing to provide real-time structural assessment. To overcome these challenges, an automated damage detection system is developed that leverages the superior feature extraction and pattern recognition capabilities of deep learning for image processing and data analysis. The method enables accurate identification of structural anomalies and deterioration patterns, demonstrating significant improvements in both cost-effectiveness and inspection efficiency (35% faster than manual methods) compared to traditional approaches. The proposed framework offers a transformative solution for intelligent infrastructure monitoring, with potential applications in preventive maintenance and safety assurance.
The power construction industry’s growth demands efficient monitoring of high-risk worker behaviors, yet traditional methods are inefficient and existing models face false alarms in complex scenes. This study proposes DSR-YOLOv8, an improved YOLOv8 algorithm integrating three modules: (1) DSRAB using deep separable convolution and global pooling to enhance subtle action features and denoising; (2) SD_SPPF with multi-scale dilated kernels to expand the receptive field while reducing computational costs; (3) dynamic region-processing with partial convolutional heads to focus on critical areas and suppress interference. Evaluated on a self-built Dangerous Behavior Dataset (DBD) containing “helmet-wearing,” “no helmet,” and “smoking” scenarios, DSR-YOLOv8 achieved 91.2% accuracy (+3.5%) and 89.7% mAP (+3.6%) over baselines, demonstrating efficient hazardous behavior detection for enhanced safety in power construction.
This paper studied the spatial characteristics of tourism economic differences in the context of sustainable environment. This paper first introduced the impact of agricultural waste pollution on the environment and economic development, as well as the spatial characteristics and influencing factors of tourism economic differences that do not consider environmental factors. Then, taking nine provinces as the research object, this paper studied the efficiency driving mechanism of tourism industry and the spatial characteristics of tourism economic differences under the sustainable environment. The outcomes of the experiment demonstrated that environmental contamination would lower the tourism sector's economic efficiency. The study findings indicate that there was a 0.011 drop in the Theil index of tourist industry efficiency, a narrowing of the regional difference in tourism green development level, and a progressive rise in the regional efficiency difference as a significant contributing factor to the total regional difference.
In view of the fact that traditional design relies on experience and static simulation software, with which it is difficult to meet the needs of dynamic scenes and multiple users, and the lighting energy consumption accounts for a high proportion of total building energy consumption, this paper constructs an intelligent, algorithm-driven indoor lighting design optimization scheme. The scheme integrated architectural space characteristics, environmental dynamic data, and user behavior patterns and realized the collaborative optimization of the lighting among multiple objectives, such as energy consumption, comfort, and functionality, through an innovative three-tiered architecture model. The empirical analysis showed that the energy consumption of lighting systems optimized by intelligent algorithms was significantly reduced in different functional areas, with an average reduction of over 15%, and the user comfort and health indicators were also greatly improved. This scheme provides a new direction for indoor lighting design, promotes the development of the intelligent building industry, and helps popularize green buildings.
Vocal music performance is a unique art form that demands both refined singing skills and expressive stage presence. Many non-professional performers lack the stage experience needed to leave a lasting impression on their audiences, especially when compared with performers from music schools or professional groups. This deficiency contributes to employment difficulties for vocal performers. To address this issue, the study proposed using distributed edge computing to thoroughly evaluate the effectiveness of vocal performance, with validation conducted through simulation experiments. This research sought to improve performance evaluation, foster cultural development, enhance public aesthetic appreciation, and strengthen cultural services, thereby advancing national cultural initiatives.
Landscape designers play a crucial role in enhancing the quality of living spaces. Traditional design methods, centered around designer experience, are time-consuming and costly, with limitations in design diversity. This paper presents an automatic generation and beautification model for landscape design based on deep learning to improve efficiency and overcome aesthetic limitations. Experimental results show that NLP technology within the model enhances text processing capabilities and reduces data processing errors, while the residual module CNN improves the quality of auto-generated models. Incorporating point cloud technology enhances the depiction of local details in 3D models. Furthermore, the model supports design optimization and style conversion based on textual descriptions, offering diversified design elements for landscape design automation.
This paper presents an intelligent indoor lighting control system based on deep learning. The system employs a convolutional neural network to optimize the layout and positioning accuracy of indoor visible light communication sources and integrates a long short-term memory network with a backpropagation neural network to build a smart lighting prediction module. Experimental results demonstrate that the proposed system reduces indoor lighting parameter failure rates to below 12%, expands the effective area of signal-to-noise ratio, and lowers personnel positioning error by up to 18%. Furthermore, the model achieves high prediction accuracy when trained on historical lighting behavior data, with predicted lighting states closely matching actual user preferences. These improvements enhance user comfort and enable more personalized and energy-efficient lighting control.
Due to the widespread use of the industrial internet of things, the industrial control system has steadily transformed into an intelligent and informational one. To increase the industrial control system's security, based on industrial control system assets, this paper provides a method of threat modeling, attributing, and reasoning. First, this method characterizes the asset threat of an industrial control system by constructing an asset security ontology based on the asset structure. Second, this approach makes use of machine learning to identify assets and attribute the attacker's attack path. Subsequently, inference rules are devised to replicate the attacker's attack path, thereby reducing the response time of security personnel to threats and strengthening the semantic relationship between asset security within industrial control systems. Finally, the process is used in the simulation environment and real case scenario based on the power grid, where the assets and attacks are mapped. The actual attack path is deduced, and it demonstrates the approach's effectiveness.
In recent years, air pollution impacts athletes during outdoor training, especially in endurance and explosive-power training. This paper uses the boundary image inverse optimization algorithm and dynamic scene three-dimensional analysis model to study this impact. First, it briefly introduces the algorithms' application in sports training and existing research. Then, by combining dynamic scene classification technology, it quantifies the training process and proposes an improved boundary image inverse optimization algorithm. Finally, it uses the 3D analysis model to classify the training process and verifies it with data under different air-pollution levels. The study shows that the same-level air pollution affects outdoor endurance and explosive-power training differently. The proposed model can effectively quantify pollution's impact, offer training suggestions, improve training effects, and reduce pollution-related harm to athletes.
Most of the collision-related decisions of ships at sea depend on the working experience of drivers and determining a reasonable avoidance decision quickly when facing a multivessel encounter situation is difficult, so applying intelligent algorithms to assist these decisions is necessary. On the basis of this, the authors researched the construction of intelligent decision support systems for ship collision avoidance that relies on an anthropomorphic physics optimization algorithm. They used this algorithm to obtain the global range optimal solutions through iteration, which provides effective decisions for ship collision avoidance. The experiments were designed to simulate and analyze the ship collision avoidance decision model. The results showed that the decision-making system based on the anthropomorphic physics optimization algorithm can provide an effective collision avoidance decision scheme.
The development of the modern logistics industry in China is relatively backward, and the logistics industry has always shown characteristics of high emissions, high energy consumption, and high pollution. The low-carbon transformation and development of the logistics industry has become a prominent issue. At present, research on low-carbon logistics mainly focuses on its definition and characteristics. The research on its influencing factors is mostly qualitative analysis, and there is relatively little research on performance evaluation. This article establishes a performance evaluation index system from the perspective of low-carbon environmental protection, including information technology level, warehousing, transportation, packaging, etc., and constructs a comprehensive evaluation model that combines an analytic hierarchy process with data envelopment analysis for low-carbon logistics performance evaluation. By reviewing the collected data and taking five logistics enterprises in China as examples, the feasibility of low-carbon logistics performance evaluation indicators and models was verified.
Traditional culture is gradually being forgotten in the process of modernization, leading to insufficient application of traditional elements in movies and animations, which in turn affects their artistic value and market performance. To avoid the dilemma of high and low development, Chinese film and animation production needs to find a healthier and more benign development path. Explored in this article is the application of traditional elements in AI driven film and television animations. To achieve this goal, in-depth research on element matching problems using a tensor canonical polyadic decomposition (CPD) model, sampling algorithm based on composite star structure, and maximum filtering method was conducted. Through these algorithms and models, a way to effectively utilize traditional elements in AI driven movies and TV animations can be found. This article provides new ideas and directions for the development of China's film and animation industry and offers new possibilities for the application of AI technology in the cultural realm.
Global swimming competitiveness, driven by enhanced training science, necessitates innovative strategies. High-level swimmers worldwide follow common principles, underscoring AI's potential in training enhancement. Software engineering education actively seeks to align talent development with industry needs, particularly via AI-driven creation of immersive training environments for students. Despite UWB's prominence among wireless positioning technologies and ongoing Chinese research, its application in sports arenas is yet limited. This paper presents a graphical representation of AI in swimming training, proposing a 20% accuracy boost and data visualization, enriching discussions on AI-assisted learning platforms in software engineering and athletics.
How to improve marketing efficiency and how to formulate scientific, effective and practical marketing strategies according to changes in the market environment are important issues facing current smart home companies. This paper adopts the analysis methods such as PEST macro model analysis method, Porter's five forces model, SWOT analysis model and STP analysis, taking Suzhou as the research object, conducts an in-depth analysis of the development environment of smart home, and clarifies the suitable smart home market. From the aspects of 6P marketing mix (product, price, channel, promotion, public relations and political rights), it proposes a Suzhou smart home marketing mix strategy based on mid-to-high-end smart home products. The research results of this paper will promote the development of the city's smart home market and have important reference value for other cities to carry out smart home marketing and can also be used as a reference for relevant government departments to formulate policies.
Due to the widespread adoption of social networks, image-text comments have become a prevalent mode of emotional expression compared to traditional text descriptions. However, there are currently two major challenges. The first is the question of how to extract rich representations effectively from both text and images, and the second is the question of how to extract cross-modal shared emotion features. This study proposes a multimodal sentiment analysis method based on a deep feature interaction network (DFINet). It leverages word-to-word graphs and deep attention interaction networks (DAIN) to learn text representations effectively from multiple subspaces. Additionally, it introduces a cross-modal attention interaction network to extract cross-modal shared emotion features efficiently. This approach helps alleviate the difficulties associated with acquiring image-text features and representing cross-modal shared emotion features. Experimental results on the Yelp dataset demonstrate the effectiveness of the DFINet method.
