Human motion modeling is a core technology in computer animation,game development,and humancomputer interaction.In particular,generating natural and coherent in-between motion using only the initial and terminal frame...Human motion modeling is a core technology in computer animation,game development,and humancomputer interaction.In particular,generating natural and coherent in-between motion using only the initial and terminal frames remains a fundamental yet unresolved challenge.Existing methods typically rely on dense keyframe inputs or complex prior structures,making it difficult to balance motion quality and plausibility under conditions such as sparse constraints,long-term dependencies,and diverse motion styles.To address this,we propose a motion generation framework based on a frequency-domain diffusion model,which aims to better model complex motion distributions and enhance generation stability under sparse conditions.Our method maps motion sequences to the frequency domain via the Discrete Cosine Transform(DCT),enabling more effective modeling of low-frequency motion structures while suppressing high-frequency noise.A denoising network based on self-attention is introduced to capture long-range temporal dependencies and improve global structural awareness.Additionally,a multi-objective loss function is employed to jointly optimize motion smoothness,pose diversity,and anatomical consistency,enhancing the realism and physical plausibility of the generated sequences.Comparative experiments on the Human3.6M and LaFAN1 datasets demonstrate that our method outperforms state-of-the-art approaches across multiple performance metrics,showing stronger capabilities in generating intermediate motion frames.This research offers a new perspective and methodology for human motion generation and holds promise for applications in character animation,game development,and virtual interaction.展开更多
The dynamics of phase separation in H–He binary systems within gas giants such as Jupiter and Saturn exhibit remarkable complexity, yet lack systematic investigation. Through large-scale machine-learning-accelerated ...The dynamics of phase separation in H–He binary systems within gas giants such as Jupiter and Saturn exhibit remarkable complexity, yet lack systematic investigation. Through large-scale machine-learning-accelerated molecular dynamics simulations spanning broad temperature-pressure-composition(2000–10000 K, 1–7 Mbar,pure H to pure He) regimes, we systematically determine self and mutual diffusion coefficients in H–He systems and establish a six-dimensional framework correlating temperature, pressure, helium abundance, phase separation degree, diffusion coefficients, and anisotropy. Key findings reveal that hydrogen exhibits active directional migration with pronounced diffusion anisotropy, whereas helium passively aggregates in response. While the conventional mixing rule underestimates mutual diffusion coefficients by neglecting velocity cross-correlations,the assumption of an ideal thermodynamic factor(Q = 1) overestimates them due to unaccounted non-ideal thermodynamic effects—both particularly pronounced in strongly phase-separated regimes. Notably, hydrogen's dual role, anisotropic diffusion and bond stabilization via helium doping, modulates demixing kinetics. Large-scale simulations(216,000 atoms) propose novel phase-separation paradigms, such as “hydrogen bubble/wisp” formation, challenging the classical “helium rain” scenario, striving to bridge atomic-scale dynamics to planetary-scale phase evolution.展开更多
Missing values in radionuclide diffusion datasets can undermine the predictive accuracy and robustness of the machine learning(ML)models.In this study,regression-based missing data imputation method using a light grad...Missing values in radionuclide diffusion datasets can undermine the predictive accuracy and robustness of the machine learning(ML)models.In this study,regression-based missing data imputation method using a light gradient boosting machine(LGBM)algorithm was employed to impute more than 60%of the missing data,establishing a radionuclide diffusion dataset containing 16 input features and 813 instances.The effective diffusion coefficient(D_(e))was predicted using ten ML models.The predictive accuracy of the ensemble meta-models,namely LGBM-extreme gradient boosting(XGB)and LGBM-categorical boosting(CatB),surpassed that of the other ML models,with R^(2)values of 0.94.The models were applied to predict the D_(e)values of EuEDTA^(−)and HCrO_(4)^(−)in saturated compacted bentonites at compactions ranging from 1200 to 1800 kg/m^(3),which were measured using a through-diffusion method.The generalization ability of the LGBM-XGB model surpassed that of LGB-CatB in predicting the D_(e)of HCrO_(4)^(−).Shapley additive explanations identified total porosity as the most significant influencing factor.Additionally,the partial dependence plot analysis technique yielded clearer results in the univariate correlation analysis.This study provides a regression imputation technique to refine radionuclide diffusion datasets,offering deeper insights into analyzing the diffusion mechanism of radionuclides and supporting the safety assessment of the geological disposal of high-level radioactive waste.展开更多
A medical image encryption is proposed based on the Fisher-Yates scrambling,filter diffusion and S-box substitution.First,chaotic sequence associated with the plaintext is generated by logistic-sine-cosine system,whic...A medical image encryption is proposed based on the Fisher-Yates scrambling,filter diffusion and S-box substitution.First,chaotic sequence associated with the plaintext is generated by logistic-sine-cosine system,which is used for the scrambling,substitution and diffusion processes.The three-dimensional Fisher-Yates scrambling,S-box substitution and diffusion are employed for the first round of encryption.The chaotic sequence is adopted for secondary encryption to scramble the ciphertext obtained in the first round.Then,three-dimensional filter is applied to diffusion for further useful information hiding.The key to the algorithm is generated by the combination of hash value of plaintext image and the input parameters.It improves resisting ability of plaintext attacks.The security analysis shows that the algorithm is effective and efficient.It can resist common attacks.In addition,the good diffusion effect shows that the scheme can solve the differential attacks encountered in the transmission of medical images and has positive implications for future research.展开更多
The interdiffusion coefficients in Al_(0.2)CoCrFeNi,CoCrCu_(0.2)FeNi,and CoCrFeMn_(0.2)Ni high-entropy alloys were efficiently determined by combining diffusion couple experiments and high-throughput determination of ...The interdiffusion coefficients in Al_(0.2)CoCrFeNi,CoCrCu_(0.2)FeNi,and CoCrFeMn_(0.2)Ni high-entropy alloys were efficiently determined by combining diffusion couple experiments and high-throughput determination of interdiffusion coefficients(HitDIC)software at 1273−1373 K.The results show that the addition of Al,Cu,and Mn to CoCrFeNi high-entropy alloys promotes the diffusion of Co,Cr,and Fe atoms.The comparison of tracer diffusion coefficients indicates that there is no sluggish diffusion in tracer diffusion on the thermodynamic temperature scale for the present Al_(0.2)CoCrFeNi,CoCrCu_(0.2)FeNi,and CoCrFeMn_(0.2)Ni high-entropy alloys.The linear relationship between diffusion entropy and activation energy reveals that the diffusion process of atoms is unaffected by an increase in the number of components as long as the crystal structure remains unchanged.展开更多
The application of generative artificial intelligence(AI)is bringing about notable changes in anime creation.This paper surveys recent advancements and applications of diffusion and language models in anime generation...The application of generative artificial intelligence(AI)is bringing about notable changes in anime creation.This paper surveys recent advancements and applications of diffusion and language models in anime generation,focusing on their demonstrated potential to enhance production efficiency through automation and personalization.Despite these benefits,it is crucial to acknowledge the substantial initial computational investments required for training and deploying these models.We conduct an in-depth survey of cutting-edge generative AI technologies,encompassing models such as Stable Diffusion and GPT,and appraise pivotal large-scale datasets alongside quantifiable evaluation metrics.Review of the surveyed literature indicates the achievement of considerable maturity in the capacity of AI models to synthesize high-quality,aesthetically compelling anime visual images from textual prompts,alongside discernible progress in the generation of coherent narratives.However,achieving perfect long-form consistency,mitigating artifacts like flickering in video sequences,and enabling fine-grained artistic control remain critical ongoing challenges.Building upon these advancements,research efforts have increasingly pivoted towards the synthesis of higher-dimensional content,such as video and three-dimensional assets,with recent studies demonstrating significant progress in this burgeoning field.Nevertheless,formidable challenges endure amidst these advancements.Foremost among these are the substantial computational exigencies requisite for training and deploying these sophisticated models,particularly pronounced in the realm of high-dimensional generation such as video synthesis.Additional persistent hurdles include maintaining spatial-temporal consistency across complex scenes and mitigating ethical considerations surrounding bias and the preservation of human creative autonomy.This research underscores the transformative potential and inherent complexities of AI-driven synergy within the creative industries.We posit that future research should be dedicated to the synergistic fusion of diffusion and autoregressive models,the integration of multimodal inputs,and the balanced consideration of ethical implications,particularly regarding bias and the preservation of human creative autonomy,thereby establishing a robust foundation for the advancement of anime creation and the broader landscape of AI-driven content generation.展开更多
Accurately identifying building distribution from remote sensing images with complex background information is challenging.The emergence of diffusion models has prompted the innovative idea of employing the reverse de...Accurately identifying building distribution from remote sensing images with complex background information is challenging.The emergence of diffusion models has prompted the innovative idea of employing the reverse denoising process to distill building distribution from these complex backgrounds.Building on this concept,we propose a novel framework,building extraction diffusion model(BEDiff),which meticulously refines the extraction of building footprints from remote sensing images in a stepwise fashion.Our approach begins with the design of booster guidance,a mechanism that extracts structural and semantic features from remote sensing images to serve as priors,thereby providing targeted guidance for the diffusion process.Additionally,we introduce a cross-feature fusion module(CFM)that bridges the semantic gap between different types of features,facilitating the integration of the attributes extracted by booster guidance into the diffusion process more effectively.Our proposed BEDiff marks the first application of diffusion models to the task of building extraction.Empirical evidence from extensive experiments on the Beijing building dataset demonstrates the superior performance of BEDiff,affirming its effectiveness and potential for enhancing the accuracy of building extraction in complex urban landscapes.展开更多
In this study,DyF_(3)powder was sprayed onto the polar and side surfaces of the magnets to determine the anisotropic diffusion mechanism of Dy in the sintered Nd-Fe-B magnet.The coercivity and squareness of the magnet...In this study,DyF_(3)powder was sprayed onto the polar and side surfaces of the magnets to determine the anisotropic diffusion mechanism of Dy in the sintered Nd-Fe-B magnet.The coercivity and squareness of the magnet in which the diffusion of Dy is perpendicular to the c-axis(a-magnet)are lower than those of the magnet with the diffusion of Dy parallel to the c-axis(c-magnet).Compared with the c-magnet,the a-magnet has a longer Dy-enrichment region from the diffusion surface,where Dy is enriched in the 2:14:1 grain.By contrast,the Dy concentration in the grain boundaries beyond the Dy enrichment region is lower in the a-magnet.Moreover,the Dy shells beyond the Dy enrichment region in the a-magnet are distributed on the side surfaces of the 2:14:1 grains but not on the polar surfaces.Based on the micromagnetic simulation,the Dy shells on the polar surfaces of the grains are more effective in enhancing coercivity.According to first-principle calculations,Dy migrating through 001 into the Nd vacancy in the Nd_(2)Fe_(14)B crystal has a higher diffusion barrier,thus indicating that the lattice diffusion of Dy parallel to the c-axis is more difficult.展开更多
Finding suitable initial noise that retains the original image’s information is crucial for image-to-image(I2I)translation using text-to-image(T2I)diffusion models.A common approach is to add random noise directly to...Finding suitable initial noise that retains the original image’s information is crucial for image-to-image(I2I)translation using text-to-image(T2I)diffusion models.A common approach is to add random noise directly to the original image,as in SDEdit.However,we have observed that this can result in“semantic discrepancy”issues,wherein T2I diffusion models misinterpret the semantic relationships and generate content not present in the original image.We identify that the noise introduced by SDEdit disrupts the semantic integrity of the image,leading to unintended associations between unrelated regions after U-Net upsampling.Building on the widely-used latent diffusion model,Stable Diffusion,we propose a training-free,plugand-play method to alleviate semantic discrepancy and enhance the fidelity of the translated image.By leveraging the deterministic nature of denoising diffusion implicit models(DDIMs)inversion,we correct the erroneous features and correlations from the original generative process with accurate ones from DDIM inversion.This approach alleviates semantic discrepancy and surpasses recent DDIM-inversion-based methods such as PnP with fewer priors,achieving a speedup of 11.2 times in experiments conducted on COCO,ImageNet,and ImageNet-R datasets across multiple I2I translation tasks.展开更多
Diffusion of solutes significantly affects the coarsening rate of γ'precipitates in precipitation-hardened high entropy alloys(PH-HEAs).In this work,we systematically study the refractory solutes M(Hf,Nb,Ta,Mo,W,...Diffusion of solutes significantly affects the coarsening rate of γ'precipitates in precipitation-hardened high entropy alloys(PH-HEAs).In this work,we systematically study the refractory solutes M(Hf,Nb,Ta,Mo,W,Re,Ru)diffusion in face-centered-cubic(FCC)NiCoFeCr lattice through a combination of first-principles calculations,diffusion couples,and coarsening of γ'precipitates experiments.Our calculations reveal that there exists a stronger negative correlation between solute diffusivity and Young’s modulus than between solute diffusivity and atomic size;i.e.,the higher the Young’s modulus,the more difficult solute diffusion is.Based on the electronic structure analysis,the underlying origins for such a relation-ship could be ascribed to the fact that solutes with high Young’s modulus have stronger bonds with neighboring host atoms,less compressibility,and thus poor diffusivity.Afterwards,the main interdiffu-sion coefficients of three refractory elements with similar atomic sizes and increasing Young’s modulus,Mo,W,and Re,at 1150℃in(NiCoFeCr)_(92)Al_(3)Ti_(3)M_(2)are,in order of magnitude,D_(MoMo)^(Ni)>D_(WW)^(Ni)>D_(ReRe)^(Ni),as determined by the diffusion-couple experiments.Further investigations on the coarsening kinetics of precipitates confirmed the additions of refractory elements improve the coarsening resistance of γ'pre-cipitates in the order of Re>W>Mo.The trends in the diffusivity determined by experiment and simulation are in excellent agreement.More importantly,the Young’s modulus effect for the diffusion of refractory solutes in HEAs is also carefully analyzed and discussed.Our present findings will give new insights into future design of γ'-strengthened HEAs for high-temperature structural applications.展开更多
Rechargeable aqueous zinc-metal batteries (AZMBs) are promising candidates for large-scale energy storage systems due to their low cost and high safety.However,their performance and sustainability are significantly hi...Rechargeable aqueous zinc-metal batteries (AZMBs) are promising candidates for large-scale energy storage systems due to their low cost and high safety.However,their performance and sustainability are significantly hindered by the sluggish desolvation kinetics at the electrode/electrolyte interface and the corresponding hydrogen evolution reaction where active water molecules tightly participate in the Zn(H_(2)O)_(6)^(2+)solvation shell.Herein,learnt from self-generated solid electrolyte interphase (SEI) in anodes,the dielectric but ion-conductive zinc niobate nanoparticles artificial layer is constructed on metallic Zn surface (ZNB@Zn),acting as a rapid desolvation promotor.The zincophilic and dielectric-conductive properties of ZNB layer accelerate interfacial desolvation/diffusion and suppress surface corrosion or dendrite formation,achieving uniform Zn plating/stripping behavior,as confirmed by electronic/optical microscopies and interface spectroscopical measurements together with theoretical calculations.Consequently,the as-prepared ZNB@Zn electrode exhibits excellent cycling stability of over 2000 h and robust reversibility (99.54%) even under high current density and depth of discharge conditions.Meanwhile,the assembled ZNB@Zn-based full cell displays high capacity-retention rate of 80.21%after 3000 cycles at 5 A g^(-1)and outstanding rate performance up to 10 A g^(-1).The large-areal pouch cell is stabilized for hundreds of cycles,highlighting the bright prospects of the dielectric but ion-conductive layer in further application of AZMBs.展开更多
Absorption and desorption processes of hydrogen in metals are facilitated by alloying elements;however,the formation of secondary phases often reduces storage capacity.The alloying effect on the hydrogen kinetics has ...Absorption and desorption processes of hydrogen in metals are facilitated by alloying elements;however,the formation of secondary phases often reduces storage capacity.The alloying effect on the hydrogen kinetics has been examined by time-lag permeation measurement,which lacks spatial resolution and yields the averaged diffusion coefficient from multiple phases.Here,we report an advanced scanning Kelvin probe force microscopy,combined with in-situ hydrogen loading system for submicron-scale measurement of diffusion kinetics in metals.Successive probing of the surface during hydrogen loading detects the temporal and spatial variations in the surface potential,enabling the estimation of diffusion coefficient.Not only for a single-phase magnesium but also for multiphase titaniumiron based alloys,we can obtain the diffusion coefficients of hydrogen in each phase.The estimated diffusion coefficients for TiFe alloys are higher than that for the pristine TiFe intermetallic compound,due to alloying elements that reduce the diffusion barrier and modify bond character.Our approach paves the way to the microscopic understanding of hydrogen diffusion in metals.展开更多
In this paper,we study the eigenvalue problem of the Markov diffusion operator L^(2),and give generalized inequalities for eigenvalues of the operator L^(2)on a Markov diffusion triple.By applying these inequalities,w...In this paper,we study the eigenvalue problem of the Markov diffusion operator L^(2),and give generalized inequalities for eigenvalues of the operator L^(2)on a Markov diffusion triple.By applying these inequalities,we then get some new universal bounds for eigenvalues of a special Markov diffusion operator L^(2)on bounded domains in an Euclidean space.Moreover,our results can reveal the relationship between the(k+1)-th eigenvalue and the first k eigenvalues in a relatively straightforward manner.展开更多
Despite extensive research on computational geomechanics and fluid dynamics,accurately simulating convection-diffusion(CD)processes in complex fractured systems remains a significant challenge.This study develops a 3D...Despite extensive research on computational geomechanics and fluid dynamics,accurately simulating convection-diffusion(CD)processes in complex fractured systems remains a significant challenge.This study develops a 3D numerical framework for modelling CD processes in fractured geological media.The framework integrates Darcy's law and Fick's law,considering flux interactions between the matrix and fractures.The meshing strategy generates high-quality grids even in scenarios involving intersecting fractures.Then,a unified numerical scheme for solving the CD system is proposed.The novelties of this work include:(1)The proposed framework enables effective simulation of 3D fractured media,including more complex fractured vuggy media;(2)The numerical method precisely discretizes the CD terms in governing equations;(3)A Non-Orthogonal Correction(NOC)method,combined with an adaptive time integration scheme,is proposed for eliminating errors induced by skewed grids;and(4)The effects of fracture patterns and heterogeneity on flow are thoroughly analysed.The proposed method is validated through benchmark tests,demonstrating the superiority of the NOC method compared to classical methods.Further analysis reveals the evolution characteristics of pressure and concentration,offering insights into the effects of fracture patterns and heterogeneity on flow and diffusion processes.展开更多
Gravitational potential energy （GPE） source and sink due to stirring and cabbeling associated with sigma dif fusion/ advection is analyzed. It is shown that GPE source and sink is too big, and they are not closely l...Gravitational potential energy （GPE） source and sink due to stirring and cabbeling associated with sigma dif fusion/ advection is analyzed. It is shown that GPE source and sink is too big, and they are not closely linked to physical property distribution, such as temperature, salinity and velocity. Although the most frequently quoted advantage of sigma coordinate models are their capability of dealing with topography; the exces sive amount of GPE source and sink due to stirring and cabbeling associated with sigma diffusion/advec tion diagnosed from our analysis raises a very serious question whether the way lateral diffusion/advection simulated in the sigma coordinates model is physically acceptable. GPE source and sink in three coordinates is dramatically different in their magnitude and patterns. Overall, in terms of simulating lateral eddy diffu sion and advection isopycnal coordinates is the best choice and sigma coordinates is the worst. The physical reason of the excessive GPE source and sink in sigma coordinates is further explored in details. However, even in the isopycnal coordinates, simulation based on the Eulerian coordinates can be contaminated by the numerical errors associated with the advection terms.展开更多
Grain boundary diffusion technology is pivotal in the preparation of high-performance NdFeB magnets.This study investigates the factors that affect the efficiency of grain boundary diffusion,starting from the properti...Grain boundary diffusion technology is pivotal in the preparation of high-performance NdFeB magnets.This study investigates the factors that affect the efficiency of grain boundary diffusion,starting from the properties of the diffusion matrix.Through the adjustment of the sintering process,we effectively prepared magnets with varied densities that serve as the matrix for grain boundary diffusion with TbH,diffusion.The mobility characteristics of the Nd-rich phase during the densification stage are leveraged to ensure a more extensive distribution of heavy rare earth elements within the magnets.According to the experimental results,the increase in coercivity of low-density magnets after diffusion is significantly greater than that of relatively high-density magnets.The coercivity values measured are 805.32 kA/m for low-density magnets and 470.3 kA/m for high-density magnets.Additionally,grain boundary diffusion notably enhances the density of initial low-density magnets,addressing the issue of low density during the sintering stage.Before the diffusion treatment,the Nd-rich phases primarily concentrate at the triangular grain boundaries,resulting in an increased number of cavity defects in the magnets.These cavity defects contain atoms in a higher energy state,making them more prone to transition.Consequently,the diffusion activation energy at the void defects is lower than the intracrystalline diffusion activation energy,accelerating atom diffusion.The presence of larger cavities also provides more space for atom migration,thereby promoting the diffusion process.After the diffusion treatment,the proportion of bulk Nd-rich phases significantly decreases,and they infiltrate between the grains to fill the cavity defects,forming continuous fine grain boundaries.Based on these observations,the study aims to explore how to utilize this information to develop an efficient technique for grain boundary diffusion.展开更多
Imputation of missing data has long been an important topic and an essential application for intelligent transportation systems(ITS)in the real world.As a state-of-the-art generative model,the diffusion model has prov...Imputation of missing data has long been an important topic and an essential application for intelligent transportation systems(ITS)in the real world.As a state-of-the-art generative model,the diffusion model has proven highly successful in image generation,speech generation,time series modelling etc.and now opens a new avenue for traffic data imputation.In this paper,we propose a conditional diffusion model,called the implicit-explicit diffusion model,for traffic data imputation.This model exploits both the implicit and explicit feature of the data simultaneously.More specifically,we design two types of feature extraction modules,one to capture the implicit dependencies hidden in the raw data at multiple time scales and the other to obtain the long-term temporal dependencies of the time series.This approach not only inherits the advantages of the diffusion model for estimating missing data,but also takes into account the multiscale correlation inherent in traffic data.To illustrate the performance of the model,extensive experiments are conducted on three real-world time series datasets using different missing rates.The experimental results demonstrate that the model improves imputation accuracy and generalization capability.展开更多
基金supported by the National Natural Science Foundation of China(Grant No.72161034).
文摘Human motion modeling is a core technology in computer animation,game development,and humancomputer interaction.In particular,generating natural and coherent in-between motion using only the initial and terminal frames remains a fundamental yet unresolved challenge.Existing methods typically rely on dense keyframe inputs or complex prior structures,making it difficult to balance motion quality and plausibility under conditions such as sparse constraints,long-term dependencies,and diverse motion styles.To address this,we propose a motion generation framework based on a frequency-domain diffusion model,which aims to better model complex motion distributions and enhance generation stability under sparse conditions.Our method maps motion sequences to the frequency domain via the Discrete Cosine Transform(DCT),enabling more effective modeling of low-frequency motion structures while suppressing high-frequency noise.A denoising network based on self-attention is introduced to capture long-range temporal dependencies and improve global structural awareness.Additionally,a multi-objective loss function is employed to jointly optimize motion smoothness,pose diversity,and anatomical consistency,enhancing the realism and physical plausibility of the generated sequences.Comparative experiments on the Human3.6M and LaFAN1 datasets demonstrate that our method outperforms state-of-the-art approaches across multiple performance metrics,showing stronger capabilities in generating intermediate motion frames.This research offers a new perspective and methodology for human motion generation and holds promise for applications in character animation,game development,and virtual interaction.
基金supported by the National University of Defense Technology Research Fund Projectthe National Natural Science Foundation of China under Grant Nos. 12047561 and 12104507+1 种基金the NSAF under Grant No. U1830206the Science and Technology Innovation Program of Hunan Province under Grant No. 2021RC4026。
文摘The dynamics of phase separation in H–He binary systems within gas giants such as Jupiter and Saturn exhibit remarkable complexity, yet lack systematic investigation. Through large-scale machine-learning-accelerated molecular dynamics simulations spanning broad temperature-pressure-composition(2000–10000 K, 1–7 Mbar,pure H to pure He) regimes, we systematically determine self and mutual diffusion coefficients in H–He systems and establish a six-dimensional framework correlating temperature, pressure, helium abundance, phase separation degree, diffusion coefficients, and anisotropy. Key findings reveal that hydrogen exhibits active directional migration with pronounced diffusion anisotropy, whereas helium passively aggregates in response. While the conventional mixing rule underestimates mutual diffusion coefficients by neglecting velocity cross-correlations,the assumption of an ideal thermodynamic factor(Q = 1) overestimates them due to unaccounted non-ideal thermodynamic effects—both particularly pronounced in strongly phase-separated regimes. Notably, hydrogen's dual role, anisotropic diffusion and bond stabilization via helium doping, modulates demixing kinetics. Large-scale simulations(216,000 atoms) propose novel phase-separation paradigms, such as “hydrogen bubble/wisp” formation, challenging the classical “helium rain” scenario, striving to bridge atomic-scale dynamics to planetary-scale phase evolution.
基金supported by the National Natural Science Foundation of China(No.12475340 and 12375350)Special Branch project of South Taihu Lakethe Scientific Research Fund of Zhejiang Provincial Education Department(No.Y202456326).
文摘Missing values in radionuclide diffusion datasets can undermine the predictive accuracy and robustness of the machine learning(ML)models.In this study,regression-based missing data imputation method using a light gradient boosting machine(LGBM)algorithm was employed to impute more than 60%of the missing data,establishing a radionuclide diffusion dataset containing 16 input features and 813 instances.The effective diffusion coefficient(D_(e))was predicted using ten ML models.The predictive accuracy of the ensemble meta-models,namely LGBM-extreme gradient boosting(XGB)and LGBM-categorical boosting(CatB),surpassed that of the other ML models,with R^(2)values of 0.94.The models were applied to predict the D_(e)values of EuEDTA^(−)and HCrO_(4)^(−)in saturated compacted bentonites at compactions ranging from 1200 to 1800 kg/m^(3),which were measured using a through-diffusion method.The generalization ability of the LGBM-XGB model surpassed that of LGB-CatB in predicting the D_(e)of HCrO_(4)^(−).Shapley additive explanations identified total porosity as the most significant influencing factor.Additionally,the partial dependence plot analysis technique yielded clearer results in the univariate correlation analysis.This study provides a regression imputation technique to refine radionuclide diffusion datasets,offering deeper insights into analyzing the diffusion mechanism of radionuclides and supporting the safety assessment of the geological disposal of high-level radioactive waste.
文摘A medical image encryption is proposed based on the Fisher-Yates scrambling,filter diffusion and S-box substitution.First,chaotic sequence associated with the plaintext is generated by logistic-sine-cosine system,which is used for the scrambling,substitution and diffusion processes.The three-dimensional Fisher-Yates scrambling,S-box substitution and diffusion are employed for the first round of encryption.The chaotic sequence is adopted for secondary encryption to scramble the ciphertext obtained in the first round.Then,three-dimensional filter is applied to diffusion for further useful information hiding.The key to the algorithm is generated by the combination of hash value of plaintext image and the input parameters.It improves resisting ability of plaintext attacks.The security analysis shows that the algorithm is effective and efficient.It can resist common attacks.In addition,the good diffusion effect shows that the scheme can solve the differential attacks encountered in the transmission of medical images and has positive implications for future research.
基金supported by the National Natural Science Foundation of China(No.52374372)the Natural Science Foundation of the Jiangsu Higher Education Institutions of China(No.23KJB430042)+3 种基金the Jiangsu Province Large Scientific Instruments Open Sharing Autonomous Research Filing Project,China(No.TC2023A037)the Yangzhou City−Yangzhou University Cooperation Foundation,China(No.YZ2022183)High-end Talent Support Program of Yangzhou University,China,Qinglan Project of Yangzhou University,ChinaLvyangjinfeng Talent program of Yangzhou,China.
文摘The interdiffusion coefficients in Al_(0.2)CoCrFeNi,CoCrCu_(0.2)FeNi,and CoCrFeMn_(0.2)Ni high-entropy alloys were efficiently determined by combining diffusion couple experiments and high-throughput determination of interdiffusion coefficients(HitDIC)software at 1273−1373 K.The results show that the addition of Al,Cu,and Mn to CoCrFeNi high-entropy alloys promotes the diffusion of Co,Cr,and Fe atoms.The comparison of tracer diffusion coefficients indicates that there is no sluggish diffusion in tracer diffusion on the thermodynamic temperature scale for the present Al_(0.2)CoCrFeNi,CoCrCu_(0.2)FeNi,and CoCrFeMn_(0.2)Ni high-entropy alloys.The linear relationship between diffusion entropy and activation energy reveals that the diffusion process of atoms is unaffected by an increase in the number of components as long as the crystal structure remains unchanged.
基金supported by the National Natural Science Foundation of China(Grant No.62202210).
文摘The application of generative artificial intelligence(AI)is bringing about notable changes in anime creation.This paper surveys recent advancements and applications of diffusion and language models in anime generation,focusing on their demonstrated potential to enhance production efficiency through automation and personalization.Despite these benefits,it is crucial to acknowledge the substantial initial computational investments required for training and deploying these models.We conduct an in-depth survey of cutting-edge generative AI technologies,encompassing models such as Stable Diffusion and GPT,and appraise pivotal large-scale datasets alongside quantifiable evaluation metrics.Review of the surveyed literature indicates the achievement of considerable maturity in the capacity of AI models to synthesize high-quality,aesthetically compelling anime visual images from textual prompts,alongside discernible progress in the generation of coherent narratives.However,achieving perfect long-form consistency,mitigating artifacts like flickering in video sequences,and enabling fine-grained artistic control remain critical ongoing challenges.Building upon these advancements,research efforts have increasingly pivoted towards the synthesis of higher-dimensional content,such as video and three-dimensional assets,with recent studies demonstrating significant progress in this burgeoning field.Nevertheless,formidable challenges endure amidst these advancements.Foremost among these are the substantial computational exigencies requisite for training and deploying these sophisticated models,particularly pronounced in the realm of high-dimensional generation such as video synthesis.Additional persistent hurdles include maintaining spatial-temporal consistency across complex scenes and mitigating ethical considerations surrounding bias and the preservation of human creative autonomy.This research underscores the transformative potential and inherent complexities of AI-driven synergy within the creative industries.We posit that future research should be dedicated to the synergistic fusion of diffusion and autoregressive models,the integration of multimodal inputs,and the balanced consideration of ethical implications,particularly regarding bias and the preservation of human creative autonomy,thereby establishing a robust foundation for the advancement of anime creation and the broader landscape of AI-driven content generation.
基金supported by the National Natural Science Foundation of China(Nos.61906168,62202429 and 62272267)the Zhejiang Provincial Natural Science Foundation of China(No.LY23F020023)the Construction of Hubei Provincial Key Laboratory for Intelligent Visual Monitoring of Hydropower Projects(No.2022SDSJ01)。
文摘Accurately identifying building distribution from remote sensing images with complex background information is challenging.The emergence of diffusion models has prompted the innovative idea of employing the reverse denoising process to distill building distribution from these complex backgrounds.Building on this concept,we propose a novel framework,building extraction diffusion model(BEDiff),which meticulously refines the extraction of building footprints from remote sensing images in a stepwise fashion.Our approach begins with the design of booster guidance,a mechanism that extracts structural and semantic features from remote sensing images to serve as priors,thereby providing targeted guidance for the diffusion process.Additionally,we introduce a cross-feature fusion module(CFM)that bridges the semantic gap between different types of features,facilitating the integration of the attributes extracted by booster guidance into the diffusion process more effectively.Our proposed BEDiff marks the first application of diffusion models to the task of building extraction.Empirical evidence from extensive experiments on the Beijing building dataset demonstrates the superior performance of BEDiff,affirming its effectiveness and potential for enhancing the accuracy of building extraction in complex urban landscapes.
基金supported by the National Natural Science Foundation of China(52361033)National Key Research and Development Program(2022YFB3505400)+1 种基金the Main Discipline and Technology Leaders Training Plan of Jiangxi Province(2022BCJ23007)the Jiangxi Province Postgraduate Innovation Project(YC2022-S693)。
文摘In this study,DyF_(3)powder was sprayed onto the polar and side surfaces of the magnets to determine the anisotropic diffusion mechanism of Dy in the sintered Nd-Fe-B magnet.The coercivity and squareness of the magnet in which the diffusion of Dy is perpendicular to the c-axis(a-magnet)are lower than those of the magnet with the diffusion of Dy parallel to the c-axis(c-magnet).Compared with the c-magnet,the a-magnet has a longer Dy-enrichment region from the diffusion surface,where Dy is enriched in the 2:14:1 grain.By contrast,the Dy concentration in the grain boundaries beyond the Dy enrichment region is lower in the a-magnet.Moreover,the Dy shells beyond the Dy enrichment region in the a-magnet are distributed on the side surfaces of the 2:14:1 grains but not on the polar surfaces.Based on the micromagnetic simulation,the Dy shells on the polar surfaces of the grains are more effective in enhancing coercivity.According to first-principle calculations,Dy migrating through 001 into the Nd vacancy in the Nd_(2)Fe_(14)B crystal has a higher diffusion barrier,thus indicating that the lattice diffusion of Dy parallel to the c-axis is more difficult.
基金supported in part by the National Natural Science Foundation of China(62176059)supported by The Pennsylvania State University.
文摘Finding suitable initial noise that retains the original image’s information is crucial for image-to-image(I2I)translation using text-to-image(T2I)diffusion models.A common approach is to add random noise directly to the original image,as in SDEdit.However,we have observed that this can result in“semantic discrepancy”issues,wherein T2I diffusion models misinterpret the semantic relationships and generate content not present in the original image.We identify that the noise introduced by SDEdit disrupts the semantic integrity of the image,leading to unintended associations between unrelated regions after U-Net upsampling.Building on the widely-used latent diffusion model,Stable Diffusion,we propose a training-free,plugand-play method to alleviate semantic discrepancy and enhance the fidelity of the translated image.By leveraging the deterministic nature of denoising diffusion implicit models(DDIMs)inversion,we correct the erroneous features and correlations from the original generative process with accurate ones from DDIM inversion.This approach alleviates semantic discrepancy and surpasses recent DDIM-inversion-based methods such as PnP with fewer priors,achieving a speedup of 11.2 times in experiments conducted on COCO,ImageNet,and ImageNet-R datasets across multiple I2I translation tasks.
基金financially supported by the National Natural Science Foundation of China(No.51701061)the Natural Sci-ence Foundation of Hebei Province(No.E2019202059).
文摘Diffusion of solutes significantly affects the coarsening rate of γ'precipitates in precipitation-hardened high entropy alloys(PH-HEAs).In this work,we systematically study the refractory solutes M(Hf,Nb,Ta,Mo,W,Re,Ru)diffusion in face-centered-cubic(FCC)NiCoFeCr lattice through a combination of first-principles calculations,diffusion couples,and coarsening of γ'precipitates experiments.Our calculations reveal that there exists a stronger negative correlation between solute diffusivity and Young’s modulus than between solute diffusivity and atomic size;i.e.,the higher the Young’s modulus,the more difficult solute diffusion is.Based on the electronic structure analysis,the underlying origins for such a relation-ship could be ascribed to the fact that solutes with high Young’s modulus have stronger bonds with neighboring host atoms,less compressibility,and thus poor diffusivity.Afterwards,the main interdiffu-sion coefficients of three refractory elements with similar atomic sizes and increasing Young’s modulus,Mo,W,and Re,at 1150℃in(NiCoFeCr)_(92)Al_(3)Ti_(3)M_(2)are,in order of magnitude,D_(MoMo)^(Ni)>D_(WW)^(Ni)>D_(ReRe)^(Ni),as determined by the diffusion-couple experiments.Further investigations on the coarsening kinetics of precipitates confirmed the additions of refractory elements improve the coarsening resistance of γ'pre-cipitates in the order of Re>W>Mo.The trends in the diffusivity determined by experiment and simulation are in excellent agreement.More importantly,the Young’s modulus effect for the diffusion of refractory solutes in HEAs is also carefully analyzed and discussed.Our present findings will give new insights into future design of γ'-strengthened HEAs for high-temperature structural applications.
基金National Key R&D Program of China (2021YFA1201503)National Natural Science Foundation of China (21972164, 22279161, 12264038, 22309144)+4 种基金Natural Science Foundation of Jiangsu Province (BK. 20210130)China Postdoctoral Science Foundation (2023M733189)Jiangsu Double-Innovation PhD Program in 2022 (JSSCBS20221241)Senior Talents Fund of Jiangsu University (5501220014)fellowship funding provided by the Alexander von Humboldt Foundation。
文摘Rechargeable aqueous zinc-metal batteries (AZMBs) are promising candidates for large-scale energy storage systems due to their low cost and high safety.However,their performance and sustainability are significantly hindered by the sluggish desolvation kinetics at the electrode/electrolyte interface and the corresponding hydrogen evolution reaction where active water molecules tightly participate in the Zn(H_(2)O)_(6)^(2+)solvation shell.Herein,learnt from self-generated solid electrolyte interphase (SEI) in anodes,the dielectric but ion-conductive zinc niobate nanoparticles artificial layer is constructed on metallic Zn surface (ZNB@Zn),acting as a rapid desolvation promotor.The zincophilic and dielectric-conductive properties of ZNB layer accelerate interfacial desolvation/diffusion and suppress surface corrosion or dendrite formation,achieving uniform Zn plating/stripping behavior,as confirmed by electronic/optical microscopies and interface spectroscopical measurements together with theoretical calculations.Consequently,the as-prepared ZNB@Zn electrode exhibits excellent cycling stability of over 2000 h and robust reversibility (99.54%) even under high current density and depth of discharge conditions.Meanwhile,the assembled ZNB@Zn-based full cell displays high capacity-retention rate of 80.21%after 3000 cycles at 5 A g^(-1)and outstanding rate performance up to 10 A g^(-1).The large-areal pouch cell is stabilized for hundreds of cycles,highlighting the bright prospects of the dielectric but ion-conductive layer in further application of AZMBs.
基金supported by the Korea Institute of Science and Technology(No.2E30993).
文摘Absorption and desorption processes of hydrogen in metals are facilitated by alloying elements;however,the formation of secondary phases often reduces storage capacity.The alloying effect on the hydrogen kinetics has been examined by time-lag permeation measurement,which lacks spatial resolution and yields the averaged diffusion coefficient from multiple phases.Here,we report an advanced scanning Kelvin probe force microscopy,combined with in-situ hydrogen loading system for submicron-scale measurement of diffusion kinetics in metals.Successive probing of the surface during hydrogen loading detects the temporal and spatial variations in the surface potential,enabling the estimation of diffusion coefficient.Not only for a single-phase magnesium but also for multiphase titaniumiron based alloys,we can obtain the diffusion coefficients of hydrogen in each phase.The estimated diffusion coefficients for TiFe alloys are higher than that for the pristine TiFe intermetallic compound,due to alloying elements that reduce the diffusion barrier and modify bond character.Our approach paves the way to the microscopic understanding of hydrogen diffusion in metals.
基金Supported by the Open Research Fund of Key Laboratory of Nonlinear Analysis and Applications(Central China Normal University),Ministry of Education,P.R.China(Grant No.NAA2025ORG011)Science and Technology Plan Project of Jingmen(Grant No.2024YFZD076)+3 种基金Research Team Project of Jingchu University of Technology(Grant No.TD202006)Research Project of Jingchu University of Technology(Grant Nos.HX20240049HX20240200)the Teaching Reform Research Project of Hubei Province(Grant No.2024496)。
文摘In this paper,we study the eigenvalue problem of the Markov diffusion operator L^(2),and give generalized inequalities for eigenvalues of the operator L^(2)on a Markov diffusion triple.By applying these inequalities,we then get some new universal bounds for eigenvalues of a special Markov diffusion operator L^(2)on bounded domains in an Euclidean space.Moreover,our results can reveal the relationship between the(k+1)-th eigenvalue and the first k eigenvalues in a relatively straightforward manner.
基金financially supported by the National Natural Science Foundation of China(Grant Nos.51991392 and 42293355).
文摘Despite extensive research on computational geomechanics and fluid dynamics,accurately simulating convection-diffusion(CD)processes in complex fractured systems remains a significant challenge.This study develops a 3D numerical framework for modelling CD processes in fractured geological media.The framework integrates Darcy's law and Fick's law,considering flux interactions between the matrix and fractures.The meshing strategy generates high-quality grids even in scenarios involving intersecting fractures.Then,a unified numerical scheme for solving the CD system is proposed.The novelties of this work include:(1)The proposed framework enables effective simulation of 3D fractured media,including more complex fractured vuggy media;(2)The numerical method precisely discretizes the CD terms in governing equations;(3)A Non-Orthogonal Correction(NOC)method,combined with an adaptive time integration scheme,is proposed for eliminating errors induced by skewed grids;and(4)The effects of fracture patterns and heterogeneity on flow are thoroughly analysed.The proposed method is validated through benchmark tests,demonstrating the superiority of the NOC method compared to classical methods.Further analysis reveals the evolution characteristics of pressure and concentration,offering insights into the effects of fracture patterns and heterogeneity on flow and diffusion processes.
文摘Gravitational potential energy （GPE） source and sink due to stirring and cabbeling associated with sigma dif fusion/ advection is analyzed. It is shown that GPE source and sink is too big, and they are not closely linked to physical property distribution, such as temperature, salinity and velocity. Although the most frequently quoted advantage of sigma coordinate models are their capability of dealing with topography; the exces sive amount of GPE source and sink due to stirring and cabbeling associated with sigma diffusion/advec tion diagnosed from our analysis raises a very serious question whether the way lateral diffusion/advection simulated in the sigma coordinates model is physically acceptable. GPE source and sink in three coordinates is dramatically different in their magnitude and patterns. Overall, in terms of simulating lateral eddy diffu sion and advection isopycnal coordinates is the best choice and sigma coordinates is the worst. The physical reason of the excessive GPE source and sink in sigma coordinates is further explored in details. However, even in the isopycnal coordinates, simulation based on the Eulerian coordinates can be contaminated by the numerical errors associated with the advection terms.
基金Project supported by the National Natural Science Foundation of China(52361033)National Key Research and Development Program(2022YFB3505400)+3 种基金Ministry of Industry and Information Technology Heavy Rare Earth Special Use of Sintered NdFeB Project(TC220H06J)Academic and Technical Leaders in Major Disciplines in Jiangxi Province(2022BCJ23007)Jiangxi Province Science and Technology Cooperation Key Project(20212BDH80007)Jiangxi Graduate Student Innovation Special Fund Project(YC2023-B213)。
文摘Grain boundary diffusion technology is pivotal in the preparation of high-performance NdFeB magnets.This study investigates the factors that affect the efficiency of grain boundary diffusion,starting from the properties of the diffusion matrix.Through the adjustment of the sintering process,we effectively prepared magnets with varied densities that serve as the matrix for grain boundary diffusion with TbH,diffusion.The mobility characteristics of the Nd-rich phase during the densification stage are leveraged to ensure a more extensive distribution of heavy rare earth elements within the magnets.According to the experimental results,the increase in coercivity of low-density magnets after diffusion is significantly greater than that of relatively high-density magnets.The coercivity values measured are 805.32 kA/m for low-density magnets and 470.3 kA/m for high-density magnets.Additionally,grain boundary diffusion notably enhances the density of initial low-density magnets,addressing the issue of low density during the sintering stage.Before the diffusion treatment,the Nd-rich phases primarily concentrate at the triangular grain boundaries,resulting in an increased number of cavity defects in the magnets.These cavity defects contain atoms in a higher energy state,making them more prone to transition.Consequently,the diffusion activation energy at the void defects is lower than the intracrystalline diffusion activation energy,accelerating atom diffusion.The presence of larger cavities also provides more space for atom migration,thereby promoting the diffusion process.After the diffusion treatment,the proportion of bulk Nd-rich phases significantly decreases,and they infiltrate between the grains to fill the cavity defects,forming continuous fine grain boundaries.Based on these observations,the study aims to explore how to utilize this information to develop an efficient technique for grain boundary diffusion.
基金partially supported by the National Natural Science Foundation of China(62271485)the SDHS Science and Technology Project(HS2023B044)
文摘Imputation of missing data has long been an important topic and an essential application for intelligent transportation systems(ITS)in the real world.As a state-of-the-art generative model,the diffusion model has proven highly successful in image generation,speech generation,time series modelling etc.and now opens a new avenue for traffic data imputation.In this paper,we propose a conditional diffusion model,called the implicit-explicit diffusion model,for traffic data imputation.This model exploits both the implicit and explicit feature of the data simultaneously.More specifically,we design two types of feature extraction modules,one to capture the implicit dependencies hidden in the raw data at multiple time scales and the other to obtain the long-term temporal dependencies of the time series.This approach not only inherits the advantages of the diffusion model for estimating missing data,but also takes into account the multiscale correlation inherent in traffic data.To illustrate the performance of the model,extensive experiments are conducted on three real-world time series datasets using different missing rates.The experimental results demonstrate that the model improves imputation accuracy and generalization capability.
