Search Results (38)

Generating non-standard fonts, such as running script (e.g., XingShu), poses significant challenges due to their high stroke continuity, structural flexibility, and stylistic diversity, which traditional component-based prior knowledge methods struggle to model effectively. While diffusion models excel at capturing continuous feature spaces and stroke variations through iterative denoising, they face critical limitations: (1) style leakage, where large stylistic differences lead to inconsistent outputs due to noise interference; (2) structural distortion, caused by the absence of explicit structural guidance, resulting in broken strokes or deformed glyphs; and (3) style confusion, where similar font styles are inadequately distinguished, producing ambiguous results. To address these issues, we propose a novel skeleton-guided diffusion model with three key innovations: (1) a skeleton-constrained style rendering module that enforces semantic alignment and balanced energy constraints to amplify critical skeletal features, mitigating style leakage and ensuring stylistic consistency; (2) a cross-scale skeleton preservation module that integrates multi-scale glyph skeleton information through cross-dimensional interactions, effectively modeling macro-level layouts and micro-level stroke details to prevent structural distortions; (3) a contrastive style refinement module that leverages skeleton decomposition and recombination strategies, coupled with contrastive learning on positive and negative samples, to establish robust style representations and disambiguate similar styles. Extensive experiments on diverse font datasets demonstrate that our approach significantly improves the generation quality, achieving superior style fidelity, structural integrity, and style differentiation compared to state-of-the-art diffusion-based font generation methods. Full article

Pile–raft foundations are widely used in soft soil engineering due to their good integrity and high stiffness. However, traditional design methods independently design pile–raft foundations and superstructures, ignoring their interaction. This leads to significant deviations from actual conditions when the superstructure height increases, resulting in excessive costs and adverse effects on building stability. This study experimentally investigates the interaction characteristics of pile–raft foundations and superstructures in soft soil under different working conditions using a 1:10 geometric similarity model. The superstructure is a cast-in-place frame structure (beams, columns, and slabs) with bamboo skeletons with the same cross-sectional area as the piles and rafts, cast with concrete. The piles in the foundation use rectangular bamboo strips (side length ~0.2 cm) instead of steel bars, with M1.5 mortar replacing C30 concrete. The raft is also made of similar materials. The results show that the soil settlement significantly increases under the combined action of the pile–raft and superstructure with increasing load. The superstructure stiffness constrains foundation deformation, enhances bearing capacity, and controls differential settlement. The pile top reaction force exhibits a logarithmic relationship with the number of floors, coordinating the pile bearing performance. Designers should consider the superstructure’s constraint of the foundation deformation and strengthen the flexural capacity of inner pile tops and bottom columns for safety and economy. Full article

Within computer-aided geometric design (CAGD), Said–Ball curves are primarily adopted in domains such as 3D object skeleton modeling, vascular structure repair, and path planning, owing to their flexible geometric properties. Techniques for curve degree reduction seek to reduce computational and storage demands while striving to maintain the essential geometric attributes of the original curve. This study presents a novel degree reduction model leveraging Euclidean distance and curvature data, markedly improving the preservation of geometric features throughout the reduction process. To enhance performance further, we propose a multi-strategy enhanced coati optimization algorithm (MSECOA). This algorithm utilizes a good point set combined with opposition-based learning to refine the initial population distribution, employs a fitness–distance equilibrium approach alongside a dynamic spiral search strategy to harmonize global exploration with local exploitation, and integrates an adaptive differential evolution mechanism to boost convergence rates and robustness. Experimental results demonstrate that the MSECOA outperforms nine highly cited agorithms in terms of convergence performance, solution accuracy, and stability. The algorithm exhibits superior behavior on the IEEE CEC2017 and CEC2022 benchmark functions and demonstrates strong practical utility across four engineering optimization problems with constraints. When applied to multi-degree reduction approximation of Said–Ball curves, the algorithm’s effectiveness is substantiated through four reduction cases, highlighting its superior precision and computational efficiency, thus providing a highly effective and accurate solution for complex curve degree reduction tasks. Full article

This study proposes a GPU-accelerated Position-Based Dynamics (PBD) system for realistic and interactive cloth simulation in Extended Reality (XR) environments, and comprehensively evaluates its performance and functional capabilities on standalone XR devices, such as the Meta Quest 3. To overcome the limitations of traditional CPU-based physics simulations, we designed and optimized highly parallelized algorithms utilizing Unity’s Compute Shader framework. The proposed system achieves real-time performance by implementing efficient collision detection and response handling with complex environmental meshes (RoomMesh) and dynamic hand meshes (HandMesh), as well as capsule colliders based on hand skeleton tracking (OVRSkeleton). Performance evaluations were conducted for both single-sided and double-sided cloth configurations across multiple resolutions. At a 32 × 32 resolution, both configurations maintained stable frame rates of approximately 72 FPS. At a 64 × 64 resolution, the single-sided cloth achieved around 65 FPS, while the double-sided configuration recorded approximately 40 FPS, demonstrating scalable quality adaptation depending on application requirements. Functionally, the GPU-PBD system significantly surpasses Unity’s built-in Cloth component by supporting double-sided cloth rendering, fine-grained constraint control, complex mesh-based collision handling, and real-time interaction with both hand meshes and capsule colliders. These capabilities enable immersive and physically plausible XR experiences, including natural cloth draping, grasping, and deformation behaviors during user interactions. The technical advantages of the proposed system suggest strong applicability in various XR fields, such as virtual clothing fitting, medical training simulations, educational content, and interactive art installations. Future work will focus on extending the framework to general deformable body simulation, incorporating advanced material modeling, self-collision response, and dynamic cutting simulation, thereby enhancing both realism and scalability in XR environments. Full article

Skeleton-based human action recognition has garnered significant attention for its robustness to background noise and illumination variations. However, existing methods relying on Graph Convolutional Networks (GCNs) and Transformers exhibit inherent limitations: GCNs struggle to model interactions between non-adjacent joints due to predefined skeletal topology, while Transformers accumulate noise through unrestricted global dependency modeling. To address these challenges, we propose a Dynamic Graph Attention Network (DGAN) that dynamically integrates local structural features and global spatiotemporal dependencies. DGAN employs a masked attention mechanism to adaptively adjust node connectivity, forming a dynamic adjacency matrix that extends beyond physical skeletal constraints by selectively incorporating highly correlated joints. Additionally, a node-partition bias strategy is introduced to prioritize attention on collaboratively moving body parts, thereby enhancing discriminative feature extraction. Extensive experiments on the NTU RGB+D 60 and NTU RGB+D 120 datasets validate the effectiveness of DGAN, which outperforms state-of-the-art methods by achieving a balance between local topology preservation and global interaction modeling. Our approach provides a robust framework for skeleton-driven action recognition, demonstrating superior generalization across diverse scenarios. Full article

Slope units serve as fundamental spatial units for surface morphology modeling and multidisciplinary coupling analysis, holding significant theoretical value and practical implications in regional stability assessments, surface process simulations, and quantitative geological engineering research. The scientific delineation of slope units must simultaneously satisfy engineering implementation requirements and adhere to the unit homogeneity principle. However, conventional delineation like the hydrological process analysis method (HPAM) exhibits critical limitations, including strong threshold dependency, a low automation level, and single-attribute optimization, thereby restricting its applicability in complex scenarios. Based on the principles of unit consistency and hydrological processes, this study integrates computer graphics algorithms with hydrological process simulation techniques to propose an automated slope unit division method coupled with computer graphics and hydrological principles (SUD-CGHP). The method employs digital elevation model (DEM) input data to construct a three-stage hierarchical framework comprising (1) terrain skeleton extraction through a morphological erosion algorithm, (2) topological relationship iteration optimization, and (3) multisource parameter coupling constraints. This framework achieves automated slope unit delineation without thresholds while enabling multi-attribute fusion optimization, effectively addressing the shortcomings of HPAM. Field validation in Yanglousi Town, Hunan Province, demonstrates that SUD-CGHP-generated slope units exhibit superior internal homogeneity in flow direction, slope aspect, and gradient compared to HPAM while maintaining complete topographic–hydrological connectivity. The research findings indicate that this method significantly enhances the scientific validity and practical applicability of slope unit delineation, providing reliable spatial analysis units for multidisciplinary surface process modeling. Full article

In the rapidly evolving field of computer vision and machine learning, 3D skeleton estimation is critical for applications such as motion analysis and human–computer interaction. While stereo cameras are commonly used to acquire 3D skeletal data, monocular RGB systems attract attention due to benefits including cost-effectiveness and simple deployment. However, persistent challenges remain in accurately inferring depth from 2D images and reconstructing 3D structures using monocular approaches. The current 2D to 3D skeleton estimation methods overly rely on deep training of datasets, while neglecting the importance of human intrinsic structure and the principles of camera imaging. To address this, this paper introduces an innovative 2D to 3D gait skeleton estimation method that leverages the Brown camera distortion model and constrained optimization. Utilizing the Azure Kinect depth camera for capturing gait video, the Azure Kinect Body Tracking SDK was employed to effectively extract 2D and 3D joint positions. The camera’s distortion properties were analyzed, using the Brown camera distortion model which is suitable for this scenario, and iterative methods to compensate the distortion of 2D skeleton joints. By integrating the geometric constraints of the human skeleton, an optimization algorithm was analyzed to achieve precise 3D joint estimations. Finally, the framework was validated through comparisons between the estimated 3D joint coordinates and corresponding measurements captured by depth sensors. Experimental evaluations confirmed that this training-free approach achieved superior precision and stability compared to conventional methods. Full article

Predicting human future motion holds significant importance in the domains of autonomous driving and public safety. Kinematic features, including joint coordinates and velocity, are commonly employed in skeleton-based human motion prediction. Nevertheless, most existing approaches neglect the critical role of dynamic information and tend to degrade as the prediction length increases. To address the related constraints due to single-scale and fixed-joint topological relationships, this study proposes a novel method that incorporates joint torques estimated via Lagrangian equations as dynamic features of the human body. Specifically, the human skeleton is modeled as a multi-rigid body system, with generalized joint torques calculated based on the Lagrangian formula. Furthermore, to extract both kinematic and dynamic joint information effectively for predicting long-term human motion, we propose a Multiscale Mixed-Graph Neural Network (MS-MGNN). MS-MGNN can extract kinematic and dynamic joint features across three distinct scales: joints, limbs, and body parts. The extraction of joint features at each scale is facilitated by a single-scale mixed-graph convolution module. And to effectively integrate the extracted kinematic and dynamic features, a KD-fused Graph-GRU (Kinematic and Dynamics Fused Graph Gate Recurrent Unit) predictor is designed to fuse them. Finally, the proposed method exhibits superior motion prediction capabilities across multiple motions. In motion prediction experiments on the Human3.6 dataset, it outperforms existing approaches by decreasing the average prediction error by 9.1%, 12.2%, and 10.9% at 160 ms, 320 ms, and 400 ms for short-term prediction and 7.1% at 560 ms for long-term prediction. Full article

High-resolution phylogenetics using both morphology and molecular data reveal surfactant-like trait buffering of peripatric descendant species that facilitate resilience for supra-specific entities across geologic time. Regular polygons inscribed in circles model balanced areas of survival of various numbers of new species in one genus. This model maximizes the peripatric survival of descendant species, with populations partly in allopatric habitats and in sympatric areas. It extends the theory advanced with Willis’s Age and Area hypothesis. Hollow curves of the areas bounded between a series of inscribed regular polygons and their containing circles show a ranked progression governed by similar power laws of other phenomena, including Zipf’s law and a universal meta-law in physics. This model matches best the physics meta-law (law of laws) but is only one of several somewhat different curves generated by somewhat different processes. A rule of four can explain why most genera in vascular plants exhibit a hollow curve of optimally one to five species per genus. It implies a constraint on variation that enhances survival and provides a physics explanation for the monophyletic skeleton of macrogenera. A high-resolution form of ancestor–descendant analysis is compared to traditional phylogenetic analysis to best modeling of the demonstrable results of evolutionary processes. Arguments are advanced for the preservation of scientific concepts of taxa over cladistic clades. Full article

Focusing on the issue of the low recognition rates achieved by traditional deep-information-based action recognition algorithms, an action recognition approach was developed based on skeleton spatial–temporal and dynamic features combined with a two-stream convolutional neural network (TS-CNN). Firstly, the skeleton’s three-dimensional coordinate system was transformed to obtain coordinate information related to relative joint positions. Subsequently, this relevant joint information was encoded as a color texture map to construct the spatial–temporal feature descriptor of the skeleton. Furthermore, physical structure constraints of the human body were considered to enhance class differences. Additionally, the speed information for each joint was estimated and encoded as a color texture map to achieve the skeleton motion feature descriptor. The resulting spatial–temporal and dynamic features were further enhanced using motion saliency and morphology operators to improve their expression ability. Finally, these enhanced skeleton spatial–temporal and dynamic features were deeply fused via TS-CNN for implementing action recognition. Numerous results from experiments conducted on the publicly available datasets NTU RGB-D, Northwestern-UCLA, and UTD-MHAD demonstrate that the recognition rates achieved via the developed approach are 86.25%, 87.37%, and 93.75%, respectively, indicating that the approach can effectively improve the accuracy of action recognition in complex environments compared to state-of-the-art algorithms. Full article

Customizing and optimizing lattice materials poses a challenge to designers. This study proposed a data-driven generative method to customize and optimize lattice material. The method utilizes subdivision modeling to parametrically describe lattice morphologies and skeletons. Next, the homogenization method is employed to analyze elastic moduli for collecting a dataset. Then, a two-tiered machine learning (ML) framework is proposed to predict the elastic modulus for a forward design. The first-tier model employs polynomial regression to estimate relative density, which serves as an additional input feature for the second-tier model. The prediction accuracy of the second-tier model is improved through the additional inputs. The forward and reverse design strategies offer a flexible and accurate means of tailoring lattice properties to meet specific performance requirements. Two case studies demonstrate the practical value of the framework: customizing a lattice material to achieve a desired elastic modulus and optimizing the mechanical performance of lattice materials under relative density constraints. The results show that the prediction accuracy of the elastic modulus using the two-tiered ML model achieved an error of less than 10% compared to finite element analysis, demonstrating the reliability of the proposed approach. Furthermore, the optimization design achieved up to a 25% improvement in mechanical performance compared to conventional lattice configurations under the same relative density constraints. These findings underscore the advantages of combining generative design, machine learning, and genetic algorithms to navigate complex design spaces and achieve enhanced material performance. Full article

To address the problem of low accuracy in joint point estimation in hand pose estimation methods due to the self-similarity of fingers and easy self-obscuration of hand joints, a hand pose estimation method based on coordinate correction and graph convolution is proposed. First, the standard coordinate encoding is improved by generating an unbiased heat map, and the distribution-aware method is used for decoding coordinates to reduce the error in decoding the coordinate encoding of joints. Then, the complex dependency relationship between the joints and the relationship between pixels and joints of the hand are modeled by using graph convolution, and the feature information of the hand joints is enhanced by determining the relationship between the hand joints. Finally, the skeletal constraint loss function is used to impose constraints on the joints, and a natural and undistorted hand skeleton structure is generated. Training tests are conducted on the public gesture interaction dataset STB, and the experimental results show that the method in this paper can reduce errors in hand joint point detection and improve the estimation accuracy. Full article

Recent investigations in the upper Lower–Middle Pleistocene deposits of the Megalopolis Basin (Greece) led to the discovery of several sites/findspots with abundant faunal material. Here, we provide an updated overview including new results on the micro- and macro-mammal fauna. Important new discoveries comprise partial hippopotamus skeletons from Marathousa 1 and the new Lower Pleistocene site Choremi 6, as well as a second partial elephant skeleton from Marathousa 1, including a complete tusk and the rarely found stylohyoideum. Based on the first results from the newly collected micromammals, we discuss age constraints of the sites, and we provide biostratigraphic/biochronologic remarks on key mammal taxa for the Middle Pleistocene of Greece and southeastern Europe. The presence of mammals highly dependent on freshwater for their survival, together with temperate-adapted ones in several stratigraphic layers of the basin, including those correlated with glacial stages, when conditions were colder and/or drier, indicate the capacity of the basin to retain perennial freshwater bodies under milder climatic conditions, even during the harsher glacial periods of the European Middle Pleistocene, and further support its refugial status. Yet, the smaller dimensions of the Megalopolis hippopotamuses may represent a response to the changing environmental conditions of the epoch, not optimal for hippopotamuses. Overall, the Megalopolis Basin comprises a unique fossil record for southeastern Europe and provides valuable insights into the Middle Pleistocene terrestrial ecosystems of Europe, and hominin adaptations in particular. Full article

This article primarily delves into the automatic generation approach of the park road network. The design of the park road network not only comprehensively takes into account environmental factors like terrain, vegetation, water bodies, and buildings, but also encompasses functional factors such as road coverage and accessibility. It constitutes a relatively complex design task, and traditional design methods rely significantly on the professional proficiency of designers. Based on the park vector terrain, in combination with the graphics skeleton algorithm, this study proposes an automatic generation method of the park road network considering environmental constraints. Through the utilization of the modified Douglas–Peucker algorithm and convex hull operation, the semantic information of environmental constraints is retained, domain knowledge is integrated, the skeleton graph is optimized, and issues such as road smoothness are addressed. This method can not only generate road network schemes rapidly, scientifically, and precisely, but also furnish the requisite digital model for the quantitative evaluation of the road network. Eventually, the study quantitatively assesses the experimental results via the spatial syntax theory to substantiate the efficacy of the method. Full article

Cnidarians are among the most important diploblastic organisms, elucidating many of the early stages of Metazoan evolution. However, Cnidarian fossils from Cambrian deposits have been rarely documented, mainly due to difficulties in identifying early Cnidarian representatives. Halysites, a tabulate coral from Silurian reef systems, serves as a crucial taxon for interpreting Cambrian cnidarians. Traditionally, the biological characteristics of Halysites have been analyzed using methods limited by pretreatment requirements (destructive testing) and the chamber size capacity of relevant analytical instruments. These constraints often lead to irreversible information loss and inadequate data extraction. This means that, to date, there has been no high-resolution three-dimensional mineralization analysis of Halysites. This study aims to introduce novel, non-destructive techniques to analyze the internal structure and chemical composition of Halysites. Furthermore, it seeks to elucidate the relationship between coral organisms and biomineralization in reef settings and to compare Silurian Tabulata with putative Cambrian cnidarians. Techniques such as micro-X-ray fluorescence spectrometry (micro-XRF), micro-X-ray computed tomography (micro-CT), and scanning electron microscopy (SEM) were employed in this research. With the help of high-resolution micro-CT scanning, we identify the growth pattern of Halysites, showing both lateral and vertical development. The lateral multiple-branching growth pattern of Halysites corals is first established herein. The flaggy corallite at the initial stage of branching is also observed. The micro-XRF mapping results reveal the occurrence of septa spines for Halysites, a trait previously thought rare or absent. Additionally, the ratio of coral volume to the surrounding rock was assessed, revealing that Halysites reefs were relatively sparse (volume ratio = ~30%). The cavities between Halysites likely provided more space for other organisms (e.g., rugose corals and bryozoans) when compared to other coral reef types. Additionally, we provide a comparative analysis of post-Cambrian colonial calcareous skeletons, offering insights into the structural features and growth patterns of early skeletal metazoans across the Ediacaran–Cambrian boundary. Full article