Search Results (1,474)

Pedestrian detection is a fundamental component of Advanced Driver Assistance Systems (ADAS) and plays a key role in collision avoidance and the safety of vulnerable road users. This paper presents a structured review of pedestrian detection methodologies developed between 2000 and 2025, spanning classical vision techniques and modern deep learning architectures. We organize the review into two phases. First, we examine classical methods, including Histogram of Oriented Gradients (HOG)+Support Vector Machine (SVM), Viola–Jones, Deformable Part Models, and Integral Channel Features, which established the conceptual foundations of the field. Then, we analyze state-of-the-art deep learning architectures, categorized by detector stage (one-stage vs. two-stage), localization strategy (anchor-based vs. anchor-free), feature extraction paradigm (Convolutional Neural Network (CNN)-based vs. transformer-based), output representation (bounding box vs. instance segmentation), and computational profile (lightweight vs. heavyweight). Several design principles introduced by classical methods remain visible in modern architectures, indicating that they were not fully superseded. The review also examines publicly available benchmark datasets and compares the strengths and limitations of camera-, Light Detection And Ranging (LiDAR)-, radar-, and multi-sensor-fusion-based systems for ADAS deployment. We close by identifying six open problems for the field: adversarial robustness, real-time inference under embedded constraints, detection under adverse weather, dataset bias and demographic fairness, the deployment of Bird’s-Eye View (BEV) and unified perception on automotive hardware, and explainability for safety-critical use. Full article

Recent advances in wearable, vision-based, trajectory, physiological, and multimodal sensing technologies, together with deep learning, have enabled continuous, objective, and individualized assessment of sport performance and athlete health. Unlike prior reviews that primarily focus on a single sensing modality, sport, or algorithmic series, this review integrates wearable, vision-based, trajectory, physiological, and multimodal sensing streams with deep learning models across both performance analysis and athlete health monitoring, thereby clarifying modality-task-model relationships and translational limitations. This review synthesizes recent progress in sensor-based sports intelligence, focusing on how heterogeneous data streams are transformed into performance- and health-related decision support. The reviewed applications include athlete and ball perception, multi-object tracking, pose estimation, action recognition, trajectory and tactical analysis, training-load and fatigue monitoring, injury-risk prediction, rehabilitation monitoring, and return-to-play support. Deep learning architectures, including CNNs, LSTMs, GRUs, TCNs, Transformers, attention mechanisms, graph neural networks, and multimodal fusion models, are discussed in relation to their suitability for visual, temporal, spatial, physiological, and multisource data. This review further identifies key challenges, including data heterogeneity, annotation scarcity, limited cross-sport and cross-device generalization, real-time deployment constraints, model interpretability, privacy protection, and ethical governance. Moving forward, research efforts should focus on the development of standardized datasets, reliable multimodal data fusion strategies, self-supervised and transfer learning approaches, and deployment on edge or cloud computing platforms. Additionally, enhancing interpretability through explainable AI and implementing closed-loop, individualized monitoring systems are critical. By synthesizing advances in sensing technologies, deep learning methodologies, and real-world applications, this review aims to provide a practical reference for optimizing athletic performance, preventing injuries, guiding rehabilitation, and supporting long-term health management of athletes. Full article

Reliable wearable activity recognition requires not only a class label but also an auditable indication of whether that label is supported by historical sensor evidence. We present CC-NSIEA, a label-preserving neural-plus-rule-based class-contrast evidence audit for cross-subject wearable activity recognition. A Temporal Residual Perception Network supplies the sole activity label, posterior probabilities, and a normalized temporal embedding. A read-only Training-Subject Evidence Memory retrieves global, predicted-class, and competing-class records. A rule-based Evidence Consistency Audit combines data validity, dynamic/static motion coherence, retrieval support, and class separation. When first-round evidence is insufficient, Class-Contrast Evidence Refinement performs one deterministic contrast between the predicted class and the strongest posterior competitor; the audit cannot change the neural label. The term neuro-symbolic is used only in this restricted architectural sense: a neural predictor is coupled to explicitly represent deterministic predicates and a finite rule-based controller; the method does not perform symbolic inference, theorem proving, or knowledge-graph reasoning. On five subject-disjoint outer folds of the UCI HAR official training partition, the shared perception model achieved 90.13% accuracy and 90.55% macro-F1 across 7352 out-of-fold windows from 21 subjects. Relative to a matched dynamic deterministic controller, CC-NSIEA increased Error AUPRC from 0.423802 to 0.433057 and reduced AURC from 0.035941 to 0.035913. The 10,000-resample subject-cluster bootstrap interval for the AUPRC difference was [0.001595, 0.019547]. CC-NSIEA provides an evidence-centered complement to confidence-based reliability estimation. Full article

This article compares urban planning affecting historic buildings in Berlin and Stockholm. It examines some cases of preservation, demolition and reconstruction of historic buildings: the Hansa Quarter and the Palace of the Republic in Berlin, and Sergels torg with the House of Culture and Vällingby in Stockholm. Today, while Berlin has opted for reconstruction in several cases, Stockholm is preserving the status quo achieved by the large-scale demolitions during the 1950s and 1960s. Different historic approaches in urban planning are subsumed under the categories “architecture as wellbeing” and “the automotive city.” The policy cycle serves as a framework for a qualitative content analysis of debates on urban planning in both city councils. The article tests the hypothesis whether war destructions present in Berlin, but not in Stockholm, can explain the lack of plans for reconstruction of historic buildings in Stockholm. The examination of historic developments and current legislation on German and Swedish cultural policy and the case studies of the above-named buildings yield the result that the hypothesis is proven wrong. Instead, possible explanations for the lack of will to reconstruct in Stockholm are Swedish legal tradition since the 19th century, which provides little and weak protection to historic buildings, and the “people’s home” ideology shaping the Swedish self-perception as a modern nation. International legislation on monument protection such as the ICOMOS-ICCROM Guidance on Post-Disaster and Post-Conflict Recovery and Reconstruction (2023), which becomes ever more encompassing, will perhaps introduce a future policy change. Full article

As a major tributary of the upper Yangtze River, the Chishui River possesses distinctive ecological frameworks and serves as a critical refuge for the aquatic germplasm resources of the Yangtze Basin. Acrossocheilus yunnanensis, a dominant native fish endemic to the river’s headwaters, contributes substantially to sustaining regional ecosystem functioning. Nevertheless, no population genetic research on this species has been documented in the headwater reaches of the Chishui River to date. Here, whole-genome resequencing was performed on 140 individuals from seven geographical populations of A. yunnanensis sampled from the headwaters of the Chishui River. A total of 242,504,293 high-quality single-nucleotide polymorphism (SNP) markers were identified, with an average sequencing depth of 10×. Population genetic analyses revealed observed heterozygosity (H_O) ranging from 0.290 to 0.340, expected heterozygosity (He) from 0.303 to 0.333, and nucleotide diversity (π) spanning 8.063 × 10⁻⁴ to 1.086 × 10⁻³. Genome-wide F_ROH values were generally low across the seven populations, and only several individuals within Population A exhibited clear signs of inbreeding risk. Pairwise differentiation results indicated that the Banbanqiao (A) was highly genetically divergent from the other six populations. Further genome-wide selective sweep analyses verified that A has evolved unique local adaptive signatures in pathways related to mechanosensory perception, neuromotor integration, anaerobic metabolic remodeling and protein homeostasis regulation. This study systematically characterizes the population genetic architecture and local adaptation mechanisms of headwater A. yunnanensis, providing pivotal genetic evidence for wild germplasm conservation and scientific aquatic ecosystem management across the watershed. Full article

With the rapid development of artificial intelligence technology, the transportation, industry, and healthcare fields are undergoing an intelligent evolution. These advancements have raised higher requirements for technologies such as mobile robots, wearable intelligent agents, self-driving cars, and unmanned aerial vehicles. Compared with traditional discrete sensor architectures, highly integrated sensing systems deliver superior speed, efficiency, and reliability to satisfy the stringent requirements of emerging intelligent devices. By integrating advanced technologies such as perception, communication, and computing, the process of system intelligence is accelerating, driving us into the era of embodied intelligence. Thus, sensors are no longer merely passive data collection tools but have transformed into core components that drive the connection between perception and action. To help researchers better understand this transformation and clarify the implementation path, we summarize the key technological advancements in related fields. Firstly, we review the related technological developments, including the sensor, multi-modal perception, wireless communication, and edge computing technology. Then, we explore the limitations of traditional sensors and independent computing models, especially the trade-offs among latency, energy efficiency, and system reliability. Subsequently, we introduce innovative technologies that drive the development of embodied intelligence, covering advanced learning mechanisms such as multi-agent systems, reinforcement learning, and federated learning. Finally, we compare the typical application scenarios of the two paradigms and discuss the challenges faced by existing technologies and standardization. We also look forward to future research directions in this field. Full article

Against the background of cultural consumption and the digital communication of cultural heritage, this study examines how regional cultural resources can be translated into recognizable, understandable, and shareable visual expressions in traditional food packaging. Using Huizhou architectural cultural genes and Huangshan sesame cake packaging as the research context, AI-assisted design was employed to generate packaging stimuli. A semiotics-based structural model was then developed to explore how visual symbol perception influences purchase and dissemination intentions through cultural cognition and cultural identity. Based on 287 valid responses and PLS-SEM analysis, the results show that perceptions of architectural form, color, and AI-assisted design significantly enhanced cultural cognition; graphic symbol perception showed only a weak or marginal effect, whereas structure and style perception did not show a significant effect. Cultural cognition strengthened cultural identity, which further promoted purchase and dissemination intentions. The findings provide consumer-level evidence that AI-assisted visual translation of regional architectural heritage in traditional food packaging may support potential cultural communication at the levels of consumer perception and behavioral intention by enhancing cultural cognition and cultural identity. Full article

Urban parks are widely regarded as healthy and sustainable urban infrastructures, yet their respiratory benefits depend on the coupling of design, behaviour, and perception rather than ambient PM concentrations alone. A multi-season daytime fair-weather panel study across five urban park space types integrated in situ PM monitoring, SOPARC-based behavior mapping (9173 users), dwell-time surveys, and inhalation-rate libraries to estimate per capita inhaled doses, while an on-site survey (n = 837) assessed perceived PM and its influence on space choice. Understory and water spaces exhibited the highest PM₁₀ and TSP concentrations, whereas waterfront areas were the cleanest; winter concentrations were elevated but preserved the same space-type ranking. Sports spaces had the most intense activity profiles (61.9% moderate-to-extreme), and understory and sports spaces supported the longest stays, with little seasonal change in either intensity or duration. Consequently, per capita PM exposure was highest in sports and understory spaces and lowest in water and waterfront spaces. Spaces that attracted more users also delivered higher per capita doses, indicating an overlap between popularity and high-dose micro-environments. Perceptually, 94.9% of users rated PM as low or relatively low in water spaces, whereas squares had the highest share of “moderate or worse” ratings (26.4%); 78.1% chose locations based on perceived air quality, despite weak or even negative correlations with measured PM. These findings reveal a design-driven exposure architecture in which space configuration organizes both PM concentrations and user behavior, while misperception can steer visitors, especially in winter, toward the very park micro-environments that deliver the highest inhaled doses. This study provides evidence for exposure-aware park design and management that can reduce respiratory risk while supporting sustainable outdoor recreation and healthier urban living. Full article

Emotional design is increasingly recognised within landscape architecture, particularly in the context of sustainable and user-centred environments; however, systematic and data-driven approaches that translate users’ emotional responses into concrete design parameters remain limited. To address this gap, the aim of this study is to systematically integrate users’ emotional expectations into landscape design by applying Kansei Engineering, using ornamental pools as a case study. A semantic differential survey was conducted with 91 participants, including landscape design students and experts. The experimental stimuli were developed based on a Taguchi L8 orthogonal array, enabling the systematic evaluation of five design factors (depth, interior surface colour, surface planting, form, and motion) across eight configurations. The collected data were analysed using the Taguchi method and Analysis of Variance (ANOVA) to identify optimal design combinations and quantify the relative influence of each factor. The results reveal that surface planting is the dominant factor influencing perceptions such as captivating and legible, while motion plays a key role in shaping mental restoration. The optimal configuration, characterised by shallow depth, light colour, vegetation, natural form, and dynamic water, evoked strong positive responses including captivating, aesthetically pleasing, and satisfying. This study proposes a data-driven framework for linking emotional perception with landscape design variables, contributing to the development of more socially and psychologically sustainable, user-centred, and emotionally responsive landscape environments. Full article

Educational environments, particularly those with limited resources, require affordable mobile robots capable of combining human–robot interaction, autonomous assistance, and academic support without continuous dependence on cloud services. This work presents a low-cost ROS2-based mobile robot implemented on a Raspberry Pi 4B to provide educational assistance in Spanish within controlled classroom environments. The system integrates voice interaction, text-to-speech synthesis, YOLOv8n-based object perception, a specialized door detection model, ultrasonic and inertial sensing, differential-drive control, and a hybrid natural language processing architecture based on semantic caching, local inference, and optional cloud connectivity. Two task-dependent operating modes, education and navigation, selectively activate ROS2 nodes to reduce computational load and energy consumption. Experimental tests conducted in a university classroom evaluated speech recognition, vision models, natural language processing alternatives, sensor behavior, and battery life. The speech recognition module achieved 98% accuracy under both quiet and noisy conditions. YOLOv8n achieved an F1-score of 0.975 for common classroom objects, while the specialized door detector achieved 100% recall with 58.7% precision. The semantic cache correctly resolved recurrent academic queries in the exact-match evaluation, with an average latency of 3.8 s, reducing the need for external language models in known-question scenarios. The robot operated for 96 min in education mode and 75.6 min in navigation mode. These results demonstrate that Spanish voice interaction, reactive navigation, academic question answering, and resource-aware operation can be integrated into a single low-cost edge robotic platform for educational environments. Full article

Autonomous driving systems rely on a sophisticated pipeline of artificial intelligence models to perceive, predict, and plan in dynamic environments. This review presents a systematic analysis of the machine learning and deep learning models underpinning vehicle autonomy, spanning classical convolutional neural networks (CNNs) for object detection and semantic segmentation to recurrent and Transformer-based architectures for trajectory prediction and motion planning. It also provides a critical examination of the autonomous vehicle sensor stack, including cameras, LiDAR, radar, ultrasonics, and GNSS/IMU as data acquisition systems, highlighting modality-specific AI challenges such as monocular depth estimation, 3D point cloud processing, and radar Doppler interpretation. The evolution of perception and decision-making pipelines is reviewed, contrasting modular architectures with end-to-end learning paradigms that directly map raw sensor data to control commands, and discussing their trade-offs in interpretability, safety assurance, and robustness to rare edge cases. We further survey specialized hardware accelerators and heterogeneous automotive SoCs designed to meet stringent real-time and power constraints. Industrial strategies are compared, including multi-modal sensor fusion and vision-centric approaches based on large-scale imitation learning. Finally, we identify open challenges related to robustness under adverse conditions, domain shift, causal ambiguity, and the need for interpretable and certifiable AI in safety-critical autonomous driving systems. Full article

Indoor companion, rehabilitation, logistics, laboratory transport, and service robot scenarios require mobile platforms that can follow a human operator safely and flexibly under lighting changes, occlusion, texture-poor corridors, and dynamic pedestrian environments. Vision-, LiDAR-, and UWB-based following systems can provide high perception capability, but their deployment cost, environmental dependence, and sensing complexity remain limiting factors for low-perception-dependence applications. This paper presents a passive following system for a Mecanum-wheel mobile platform based on gimbal posture perception and orthogonal odometry fusion. A rope-tensioned two-axis gimbal is mounted above a 300 mm × 300 mm × 150 mm omnidirectional chassis, and a six-axis inertial sensor installed at the top of the gimbal detects pitch and roll changes induced by user traction. A piecewise posture-to-velocity mapping model with a dead zone, saturation, low-pass filtering, and acceleration limiting converts the user’s traction intention into planar velocity commands in the vehicle coordinate frame. To reduce pose errors caused by Mecanum-wheel slip and discontinuous roller-ground contact, two orthogonal passive odometry wheels and inertial attitude estimation are fused to provide planar position feedback for closed-loop following. A prototype was implemented using an Infineon TRAVEO CYT4BB77 controller, TI DRV8701E motor drivers, six-axis IMUs, magnetic encoders, and an embedded display interface. Experiments evaluated attitude estimation accuracy, planar localization accuracy, passive following performance, gyroscope compensation, and open-loop/closed-loop following. The compensated attitude module achieved a static yaw drift of 0.45 deg/h and a dynamic attitude RMSE below 0.56 deg. Orthogonal odometry fusion produced an average positioning error of 3.8 mm over a 3000 mm linear displacement, reducing error by approximately 84.6% compared with pure Mecanum-wheel drive odometry. In a 5000 mm forward traction task, closed-loop following reduced the average distance error from 38.6 mm to 11.5 mm compared with open-loop attitude mapping. The results indicate that the proposed gimbal-orthogonal odometry architecture provides a compact, intuitive, and environment-robust solution for passive following on omnidirectional mobile platforms. Full article

Industrial heritage parks combine heritage conservation with urban public use, making it necessary to assess renewal outcomes in relation to post-construction user experiences and conditions of use. Existing studies have primarily focused on conservation and renewal practices, whereas differences in user perceptions across different types of industrial heritage parks remain relatively underexplored. This study examines Huochetou Cultural Park, Yuxiang Spinning Mill Park, and Zhushantang Industrial Heritage Park in Changsha and develops a user-perception-based post-occupancy evaluation framework. The framework comprises six dimensions: architectural space, public space, functional facilities, ecological environment, cultural value, and public participation. The Analytic Hierarchy Process (AHP) is used to determine indicator weights, and Fuzzy Comprehensive Evaluation (FCE) is applied to compare perceived post-occupancy performance, with supplementary statistical tests used to examine factor-level differences. The results show that all three parks are rated as “Good” overall, although their evaluation profiles differ. Huochetou Cultural Park performs relatively well in public space and functional facilities; Zhushantang Industrial Heritage Park shows relative strengths in architectural space and cultural value; and Yuxiang Spinning Mill Park performs relatively well in ecological environment and public participation. Ecological environment remains comparatively weak across the three parks, while public participation also warrants further attention. These findings suggest that similar overall ratings may obscure case-specific post-occupancy conditions. Accordingly, targeted recommendations are proposed for each park. This study provides a case-based reference for the post-occupancy evaluation and differentiated renewal of similar industrial heritage parks. Full article

This paper examines how technological innovation in next-generation probiotics shapes consumer trust through the lens of perceived safety. Rapid advances—spanning conventional cultures (Tier 1), postbiotics (Tier 2), and engineered microbial strains (Tier 3)—are transforming functional food architectures, yet consumer trust remains a critical determinant of their successful development, application, and adoption. Drawing on interdisciplinary evidence from food microbiology, consumer perception research, and regulatory analysis, this study examines and evaluates how these distinct technological innovation tiers alter public risk dynamics. Findings indicate that processing methodologies, media framing, and the spread of misinformation significantly influence public perceptions of microbial legitimacy, while the “Animation Gap” and “Contamination Anxiety” introduce qualitatively new cognitive friction points. Furthermore, regulatory inconsistencies across jurisdictions and variability in health claim substantiation further complicate market uptake. Streamlined case-based evidence highlights physical stability, sensory performance, and explicit value metrics that determine whether technological innovations are trusted or rejected by consumers. The paper argues that bridging the gap between scientific innovation and public acceptance requires proactive communication strategies, ethical marketing practices, and participatory engagement strategies grounded in empirical integrity. In addition, digital ecosystems, including social media and algorithm-driven content exposure, play an increasingly influential role in amplifying technology neophobia, underscoring the need for robust, targeted, evidence-based public communication in the evolving landscape of probiotic and functional food innovation. Full article

Urban branding has become an influential mechanism through which cities construct identity, shape public perception, communicate cultural distinctiveness, and guide urban transformation and place-based development. In heritage contexts, its significance extends beyond promotion by supporting the continuity, recognition, and positioning of historic places. Yet existing research has focused on formal heritage districts and visual representation, offering limited explanation of how lived historic environments sustain identity and develop a foundation for heritage-based urban branding through locally embedded socio-spatial practices. This study examines how the historic area surrounding the Mosque of Sultan al-Ashraf Qaytbay within Cairo’s City of the Dead maintains a coherent heritage identity through the interaction of architectural assets, craft production, market exchange, adaptive reuse, cultural activities, place perception, and everyday community practices. It develops a domain-based analytical framework comprising five interrelated domains: heritage asset readiness, cultural activation, place perception, emergent branding outputs, and governance and institutional mediation. The framework is applied through an interpretive spatial-observational case study based on repeated site visits, structured observation, spatial mapping, and photographic documentation. Findings show that craft production, everyday exchange, adaptive reuse, and community-based activities sustain heritage identity, collective memory, and experiential continuity. Workshops and bazaars form an interconnected production–exchange system, while galleries and cultural spaces strengthen interpretation and public engagement. However, fragmented digital visibility, weak narrative coordination, and limited institutional mediation constrain the translation of these assets into coherent branding outcomes. The study therefore distinguishes heritage identity from branding formation and offers a qualitative diagnostic framework for identifying domain alignment and misalignment, supporting context-sensitive approaches to urban transformation, heritage management, and place-based development. Full article