Search Results (506)

Convolutional neural networks (CNNs) are widely used for biomedical image classification, yet it remains unclear under which conditions training on reduced subsets of available data can provide reliable guidance during model development, how much training data is required to achieve stable and comparable performance across CNN architectures, and whether increasing the training set size always leads to improved generalization or can sometimes result in degraded performance. We study this question across four biomedical datasets (breast mammography, pediatric chest X-ray, brain tumor MRI, and skin lesion dermoscopy) using the full grid of 39 CNN architectures (1–3 convolutional layers, 16/32/64 filters) from our companion architectural study, training each configuration from scratch on seven proportions of the training data (5%, 10%, 20%, 40%, 60%, 80%, and 100%) over 5 independent runs per configuration, with the test set held at a fixed size across all sample-size conditions to ensure a like-for-like comparison of generalization performance. The analysis investigates three complementary aspects of sample-size sensitivity: the stabilization of the training–test generalization gap as training-set size increases, the reliability of architecture rankings obtained from reduced training fractions as a proxy for the full-dataset ranking, and the monotonicity of test performance with respect to training-set size. Taken together, the results point to a rough four-band pattern of reliability across the sampled fractions—unstable below $20\%$ of the training set, of uncertain overfitting status between $20\%$ and $60\%$, comparatively stable between $60\%$ and $80\%$, and potentially counterproductive beyond $80\%$—while showing that this pattern is itself dataset-dependent and offers no guarantee on architecture ranking, arguing against reduced-fraction screening as a reliable shortcut for CNN architecture selection in biomedical imaging. All code and datasets are publicly released for reproducibility. Full article

As urban renewal shifts from expansion-led development toward quality improvement of existing urban areas, the restorative quality of urban village public spaces has become an important concern in high-density human settlements. Existing restorative environment research has focused mainly on parks, green spaces, waterfronts, and other settings with strong natural attributes, while audio–visual studies have often examined environmental perception, restorative appraisal, or physiological response as separate analytical components. Limited evidence therefore exists on how scene-level visual and acoustic characteristics, subjective sensory appraisal, perceived restoration, and short-term physiological responses are related within the same analytical framework in high-density urban village public spaces. This study investigated six selected public spaces in Shenzhen urban villages using a VR-based repeated-measures experiment involving 33 participants and 198 participant–scene observations, together with computer vision indicators, psychoacoustic analysis, subjective evaluation, EDA and HRV monitoring, linear mixed-effects models, and multilevel models of statistical indirect associations. Green view index (B = 0.322, p < 0.001) and color complexity (B = 0.422, p < 0.001) were positively associated with perceived restoration, whereas building enclosure (B = −0.271, p < 0.001) and sound roughness (B = −0.328, p < 0.001) were negatively associated with perceived restoration. Green view index showed a positive indirect association with perceived restoration through visual perception (ab = 0.386, 95% CI [0.304, 0.484]), whereas roughness showed a negative indirect association through soundscape perception (ab = −0.277, 95% CI [−0.385, −0.176]). In the full six-scene models, negative objective interaction estimates were compatible with acoustic constraints on favorable visual associations, but several interaction coefficients were sensitive to scene omission and should be regarded as exploratory; at the subjective level, soundscape perception and visual perception showed a positive interaction (B = 0.095, p = 0.028). Leave-one-scene-out analysis showed that several scene-level main-effect and objective interaction estimates were sensitive to the omission of S5 or S2, whereas the directions of the green-view-index and roughness indirect associations were retained. Perceived restoration was positively correlated with the reversed EDA recovery score (r = 0.416, p < 0.001), whereas HRV showed a negative association with LAeq (B = −0.150, p = 0.008), suggesting different short-term response patterns across subjective and physiological measures. The findings suggest that restorative responses in the examined urban village scenes were associated with both subjective sensory appraisal and audio–visual interactions. The study contributes by integrating scene-level visual and acoustic characteristics, subjective sensory appraisal, perceived restoration, and short-term physiological responses within a high-density urban village context. For renewal practice, the results support the coordinated consideration of adverse sound sources, visible greenness, and visual order; however, the observed relationships should be interpreted as context-specific associations rather than as universal causal mechanisms or validated design thresholds. Full article

Gravity-matching navigation is one of the important means for the covert navigation of underwater vehicles. The production and application of gravity reference maps as core key technologies have a very significant impact on the accuracy of underwater navigation. Firstly, this paper adopts the factor analysis method to obtain the comprehensive results of nine characteristic parameters such as the standard deviation and roughness of the gravity reference map by setting a range sliding window. Secondly, the TERCOM algorithm is introduced to conduct simulation verification calculations within the sliding window. After comparing and verifying with the comprehensive results of the factor analysis characteristic parameters, the limitations of statistical methods in the evaluation of the adaptability of gravity reference maps are analyzed. Thirdly, intelligent processing methods such as the learning vector quantization neural network algorithm and the extreme learning machine are proposed. The characteristic parameters of some sliding window gravity reference maps and the simulation verification results of the TERCOM algorithm are used as training samples to predict the adaptability evaluation effect of underwater gravity navigation for other sliding windows. The results show that the prediction results are generally in good agreement with the simulation verification results of the TERCOM algorithm. Compared with the learning vector quantization neural network algorithm, the extreme learning machine algorithm exhibits superior performance in terms of classification accuracy and computational efficiency. Full article

To address dimensional reduction for continuous single-label and multi-label data, this paper proposes an improved variable precision rough set method based on maximal consistent blocks. We formulate dimensional reduction as an attribute reduction problem in continuous decision information systems, construct a distance-based tolerance relation, and design a maximal consistent block generation algorithm based on pivoted Bron–Kerbosch maximal clique mining for direct continuous data modeling. We establish a generalized variable precision rough set model, define $\beta$-approximation sets and distribution reduction objectives for single- and multi-label scenarios, analyze the stage-wise complexity of the procedure, separating polynomial stages from output-sensitive enumeration stages, and develop a discernibility matrix-based reduction algorithm. Experiments on fourteen public benchmark datasets against seven baselines under Equal-d (fixed feature number) and Nested-d (training-partition tuning) protocols show that the proposed method attains the lowest average rank under Equal-d, where the Friedman test indicates overall differences among methods and Holm-adjusted Wilcoxon comparisons confirm significant advantages over MCLS and the neighborhood rough-set dependency baseline; under Nested-d, the comparison with MCLS remains significant after Holm adjustment. Parameter sensitivity analysis, distance metric comparison, ablation study, and a resource audit further confirm the robustness and feasibility of the method. Full article

This study presents the design, embedded implementation, and screening-level experimental terrain-performance evaluation of an RHex-inspired hexapod robot using a fixed encoder-assisted alternating-tripod state-machine gait. The work aims to provide an experimentally grounded assessment of how terrain properties and practical leg-thickness variation influence the locomotion of a fabricated low-complexity legged platform. A 2 × 2 × 2 full-factorial screening design was adopted to evaluate terrain roughness, terrain compliance, and leg thickness. Four terrain conditions were tested: concrete, rocky terrain, foam mats, and grass, corresponding to smooth–rigid, rough–rigid, smooth–soft, and rough–soft surfaces, respectively. Each treatment combination was evaluated in two replicate runs using final forward displacement, lateral displacement, absolute displacement, and peak current as the response variables. The full-factorial analysis showed that terrain roughness had the clearest significant effect on forward displacement and absolute displacement, while terrain compliance significantly affected absolute displacement and showed observable trends in lateral displacement and peak current. Leg thickness did not produce a statistically significant main effect within the tested 2.5 mm and 5.0 mm configurations, fixed gait, and terrain set. The findings indicate that, for this platform and experimental scope, terrain roughness and compliance affected locomotion more strongly than the tested morphology variation. The study contributes a reproducible baseline workflow for terrain-performance evaluation in low-complexity legged robots and identifies directions for future work involving stronger replication, quantified terrain characterization, improved energy measurement, and closed-loop terrain-adaptive control. Full article

For gravitational wave observation, an optical design of the telescope has been implemented to perfectly match the laser interferometry system, and stray light analysis is used to quantify the impact of mirror roughness noise on interferometric measurement sensitivity. An off-axis six-mirror afocal optical design, comprising a parabolic primary, hyperbolic secondary and plane-parabolic collimation group, delivers an optical system with a 400 mm entrance pupil, 100× expansion ratio, and λ/30@1064 nm Root Mean Square (RMS) wavefront quality. To obtain a feasible exit pupil position which can easily integrate the laser interferometer, the theoretical mathematical relationships among the exit pupil position, primary–secondary mirror spacing, and radius of curvature of the secondary mirror and the sixth mirror are derived. Accordingly, the effective exit pupil position is extended to 174 mm to match the laser interferometer. The sixth mirror is the primary source of backscattered light. With the RMS roughness of the primary, secondary, and three folding mirrors set to 6.4 Å, 1.6 Å, and 3.7 Å, respectively, the Point Source Transmittance (PST) value can be kept below 3.6 × 10⁻¹⁰ when the RMS roughness of the sixth mirror is less than 1.1 Å. Tolerance analysis is carried out to obtain the feasible engineering distribution of optical parameters, and the statistical results show that the system RMS wavefront error is smaller than 0.01λ@1064 nm. Full article

Conventional hole finishing requires separate drilling and top/bottom chamfering, and excessive insert pressure can leave scratch-type marks on the hole wall. This study develops a clip-spring-based tool that performs drilling and top/bottom chamfering in a single machining cycle, with the chamfer depth passively set by equilibrium between the hole-wall reaction and the restoring force of a replaceable clip spring. A dual-angle insert–clip-spring interface produces a non-monotonic force drop followed by a low-incremental-stiffness plateau, separating high-force burr engagement from lower-force hole passage. Four insert-geometry and spring-bottom-shape combinations were analyzed by nonlinear finite element analysis, and the two embossed-bottom cases were supported by compression tests. The dual-angle/embossed case showed a 72% Force Dip, which compression testing reproduced together with the low-force plateau, and one-step machining confirmed process feasibility. The measured hole-wall roughness decreased fourfold, from Ra 1.7 μm to 0.4 μm. These results demonstrate a passive geometric route to Force Dip generation and self-equilibrating depth setting under the tested condition. Full article

Vehicle vibration and noise are predominantly driven by road-surface excitation, making robust prediction of these phenomena based on pavement roughness a central challenge in automotive development. This paper presents a fully automated pipeline that begins with high-resolution laser-triangulation scans of actual test tracks to produce centerline elevation profiles. These profiles are processed and classified by a MATLAB routine using ISO 8608-based power-spectral-density analysis to extract the Gh₀ roughness coefficient. Concurrently, in-service acoustic and chassis-vibration data, collected at two representative speeds, are transformed into feature vectors comprising statistical PSD descriptors. A regression model then learns the mapping from these features to Gh₀, evaluating the feasibility of mapping vehicle-borne signatures to roughness metrics. Predicted Gh₀ values drive a profile-synthesis algorithm to generate two-dimensional height grids, which are exported as CRG files and imported into a multibody simulation software (MSC ADAMS) as well as driver-in-the-loop platforms. Simulation results closely reproduce the primary excitation characteristics of the physical tracks, demonstrating a preliminary proof-of-concept pipeline for virtual road surface generation. While the cross-validated regression model indicates limited generalization on the current small dataset ($R^2=-0.2783$), the end-to-end workflow establishes the baseline integration required for future data-driven NVH simulation. To extend applicability beyond a single test vehicle, a set of Vehicle Calibration Transforms is proposed to adapt power-spectral-density features from arbitrary vehicles into the calibrated feature domain. The complete workflow promises to streamline virtual NVH validation, reduce prototype testing, and support full NVH simulator engineering in future research. Full article

Commutators of fractional integral operators play an important role in harmonic analysis due to their close connections with function regularity and partial differential equations. In this paper, we study higher-order commutators of fractional integral operators with rough variable kernels. Compared with first-order commutators, the higher-order setting involves more complicated interactions between the oscillation of the underlying BMO function and the fractional integral structure, which requires new ideas and techniques. We establish boundedness properties for these higher-order commutators under sharp conditions on the angular integrability of the variable kernels. Our results extend the existing boundedness theory of first-order commutators to higher-order cases and demonstrate the applicability of the developed techniques to a broader class of fractional integral operators with rough variable kernels. Full article

Incoloy 800H (Fe–Ni–Cr) is an iron-based superalloy that is difficult to machine because of its rapid work-hardening behaviour. Although Taguchi-based grey relational analysis has previously been applied to machining optimisation problems, studies employing an integrated grey-fuzzy framework for the dry turning of Incoloy 800H remain limited. Therefore, the present investigation aims to develop and validate a grey-fuzzy optimisation approach for determining the optimal dry turning parameters of Incoloy 800H by simultaneously minimising machining forces, surface roughness, and specific cutting pressure. Cutting speed (35, 45, and 55 m/min), feed rate (0.02, 0.04, and 0.06 mm/rev), and depth of cut (0.5, 0.75, and 1 mm) were selected as input factors, whereas feed force, thrust force, cutting force, surface roughness, and specific cutting pressure were considered output responses. Experiments were conducted using a Taguchi L27 orthogonal array (OA). The proposed methodology integrates grey relational analysis (GRA) with fuzzy logic (FL) to obtain a grey-fuzzy reasoning grade (GFRG) for multi-response optimisation. Analysis of variance (ANOVA) was employed to identify the most influential machining parameter. The results demonstrated that the grey-fuzzy approach provided a more discriminative optimisation index than conventional grey relational analysis by reducing uncertainty in multi-response decision-making. The confirmation experiments revealed an increase in GFRG from 0.550 to 0.900, corresponding to a relative improvement of 63.64% at the optimal parameter setting. The proposed methodology demonstrates that integrating grey relational analysis with fuzzy inference provides a reliable and statistically supported approach for multi-response optimisation of dry turning parameters for Incoloy 800H. Full article

This study addresses the influence of process conditions on the dimensional accuracy, geometric deviations, and surface quality of PLA parts manufactured using high-dynamics material extrusion (MEX) technology. The aim was to identify the dominant sources of variability and to assess within-condition manufacturing consistency and inter-machine consistency. The investigation considered two 3D printers, nine build locations on the working platform, two printing strategies (layer-by-layer and model-by-model), and model face orientation. Additionally, an exploratory comparison of aligned and random seam configurations and an analysis of local temperature variations within the build chamber were performed. Regular dodecahedron geometries were manufactured using a Bambu Lab P1S system and processed under identical high-quality printing parameters. Dimensional measurements were performed using a Linear 100 universal length measuring machine, while full-field geometric deviations were acquired using a GOM Scan 1 structured-light 3D scanner. Surface roughness (Ra) was measured with a MarSurf XR 20 profilometer. Part orientation is the dominant source of dimensional variability, representing the largest relative contribution to linear deviation in the mixed-effects model (ΔR² = 0.776), while local temperature variations near the printing zone were associated with location-dependent dimensional deviations. Within the supplementary temperature dataset, the regression model including temperature and printers explained 66% of the variability in mean linear dimension. This association provides indirect evidence of a thermal contribution but does not establish direct causality. The layer-by-layer strategy provided better dimensional stability than the model-by-model approach. In the exploratory seam comparison, seam configuration did not explain the orientation-dependent LD pattern. Surface roughness variability was primarily geometry-driven (ΔR² = 0.852). Variability between independent manufacturing series and specimens for linear deviation and Ra was low after accounting for the investigated factors, indicating consistent process performance under constant settings; however, the present design did not allow measurement repeatability and reproducibility to be separated. In conclusion, dimensional accuracy in high-dynamics MEX is strongly associated with part orientation, while thermal variations may represent an additional contributing factor; however, the observed correlation between thermal conditions and dimensional variability does not establish direct causality. Full article

Driving safety at tunnel portals on mountain expressways is affected by multiple factors, including driver behavior, traffic conditions, tunnel environment, and meteorological conditions. The combined effects of these factors introduce uncertainty and fuzziness into the assessment process. To address the limitations of the existing studies, which often focus on single risk factors or rely on subjective weighting methods, this study proposes an improved rough-set-based fuzzy comprehensive evaluation method incorporating the principle of minimum relative entropy. By integrating accident causation analysis, relevant standards and literature, and field investigations, a driving safety assessment indicator system for tunnel portals on mountain expressways is constructed. Attribute importance information derived from algebraic rough set theory and conditional information entropy rough set theory is first integrated. The principle of minimum relative entropy is then applied to determine the comprehensive indicator weights. Continuous and discrete indicators are then processed using membership functions and driving safety probability analysis, respectively, to establish a fuzzy comprehensive evaluation model. The results show that environmental factors and driver factors have substantial effects on driving safety at tunnel portals, with pavement icing, distracted driving, friction coefficient, and snowfall identified as the main influencing indicators. Validation using 15 tunnel portal cases on mountain expressways in Shanxi Province shows an agreement rate of 86.67% between the safety grades obtained using the proposed model and the actual accident grades. The findings provide a quantitative basis for safety risk identification, grade determination, and the operational management and control of tunnel portals on mountain expressways. Full article

In Fusion Deposition Modeling-based manufacturing, process parameters affect the mechanical and physical properties of the print. Considering these properties, accurately predicting print quality is essential. This is where machine learning (ML) models for three-dimensional (3D) print quality prediction come to the forefront. In this study, a dataset was prepared under strict operational measurement standards—utilizing the Interquartile Range (IQR) method for data sanitization—encompassing 10 material types, 2 printer types, and 4 printing parameters. Five hundred different sample combinations were prepared and printed in sets of three according to ISO 527-2 Type 4 standard dimensions. Tensile, hardness, and surface roughness tests were applied to the prepared samples. Using this validated dataset, a Hybrid Multi-Material Quality–Ensemble System–Stacking–Gradient Boosting Regressor (HMQ-ES-Stack-GBR) architecture is proposed as a diagnostic framework for multi-output quality prediction. Particularly in terms of quality outputs such as tensile strength, hardness, and surface roughness, while also providing a quantitative analysis of the effect of material type on print quality. Furthermore, a multi-objective optimization pipeline integrating three distinct meta-heuristic algorithms—Non-dominated Sorting Genetic Algorithm II (NSGA-II), Particle Swarm Optimization (PSO), and Grey Wolf Optimizer (GWO)—was coupled with the framework to systematically derive material-specific optimal processing parameter configurations. Furthermore, the study shows that open-system printers exhibit higher prediction errors than closed-system printers. Reflecting system-induced variability rather than full hardware independence. Although the study is limited to internal validation within the current experimental design and includes material imbalance across filament groups, the findings suggest that the proposed framework provides a promising diagnostic decision-support tool for pre-print quality estimation within the studied dataset. By accurately reflecting rather than physically overcoming manufacturing variability, it supports decision-making processes through pre-print quality estimation, thereby enabling proactive interventions that reduce raw material, time, and energy losses. Full article

In human–computer interaction, translating information between binary and numerical representations and human concepts remains an active research area aimed at bridging the gap while preserving the core meaning of the information. The field concerned with this transition, translation, and transformation of information is known as Human-Centric Computing (HCC). In this context, adopting a human-centered approach to information representation, such as categorization, grouping, and summarization through various forms of granulation, incorporates information into the paradigm of Granular Computing (GrC). This perspective on information processing and semantics in systems and programming has experienced continuous growth in both conceptual development and methodological interest.This paper investigates GrC research from 1990 to 2026, with a particular focus on the evolution of publications in the field. It begins with an overview of GrC, including a brief history and a discussion of its main techniques. Publications since 1990 are statistically analyzed and reviewed to identify common themes and research trends. In addition, selected research papers are examined under specific categories to provide further insights and discussion. This paper aims to highlight the evolution of GrC and identify potential future research directions. Full article

To address the issues of rough mouthfeel and monotonous flavor in plant-based beverages, this study used extruded millet flour as the raw material, with non-inoculated millet paste as the control group, and investigated the effects of six different lactic acid bacteria starter culture formulations—including two single strains (Streptococcus salivarius subsp. thermophilus and Lactobacillus delbrueckii subsp. bulgaricus) and four multi-strain consortia (2, 4, 10, and 12 strains)—on the quality of fermented millet beverages. The structural properties of the fermented millet beverages were systematically evaluated through WHC (Water Holding Capacity), LF-NMR (Low-Field Nuclear Magnetic Resonance), RVA (Rapid Visco Analyzer), rheological properties, texture, particle size, and FTIR (Fourier Transform Infrared) analyses, covering aspects such as water status, pasting behavior, viscoelasticity, textural characteristics, particle distribution, and molecular structure. In combination with volatile flavor profiling and sensory evaluation, the fermentation performance and applicability of each starter formulation were comprehensively assessed. The results showed that MFB-2 exhibited the highest viscosity and gel strength, making it suitable for thick-set products, whereas MFB-10 and MFB-12 demonstrated superior gel stability, water-holding capacity, and shelf life. Sensory evaluation further corroborated the flavor analysis, with MFB-12 showing the best aroma and overall acceptability, alongside the greatest diversity of volatile compounds, while MFB-10 presented milder acidity and favorable sensory acceptability. Collectively, MFB-10 and MFB-12 were identified as the most promising starter culture formulations for industrial-scale production of fermented millet beverages. This study provides a scientific basis for tailoring starter culture complexity to modulate the texture and flavor of plant-based fermented products. Full article