跨模态行人重识别研究的重难点主要来自于行人图像之间巨大的模态差异和模态内差异。针对这些问题,提出一种结合多尺度特征与混淆学习的网络结构。为实现高效的特征提取、缩小模态内差异,将网络设计为多尺度特征互补的形式,分别学习行...跨模态行人重识别研究的重难点主要来自于行人图像之间巨大的模态差异和模态内差异。针对这些问题,提出一种结合多尺度特征与混淆学习的网络结构。为实现高效的特征提取、缩小模态内差异,将网络设计为多尺度特征互补的形式,分别学习行人的局部细化特征与全局粗糙特征,从细粒度和粗粒度两方面来增强网络的特征表达能力。利用混淆学习策略,模糊网络的模态识别反馈,挖掘稳定且有效的模态无关属性应对模态差异,来提高特征对模态变化的鲁棒性。在大规模数据集SYSU-MM01的全搜索模式下该算法首位击中率和平均精度(mean average precision,mAP)的结果分别为76.69%和72.45%,在RegDB数据集的可见光到红外模式下该算法首位击中率和mAP的结果分别为94.62%和94.60%,优于现有的主要方法,验证了所提方法的有效性。展开更多
Tailoring grain size can improve the strength of polycrystals by regulating the proportion of grains to grain boundaries and the interaction area.As the grain size decreases to the nanoscale,the deformation mechanism ...Tailoring grain size can improve the strength of polycrystals by regulating the proportion of grains to grain boundaries and the interaction area.As the grain size decreases to the nanoscale,the deformation mechanism in polycrystals shifts from being primarily mediated by dislocations to deformation occurring within the grains and grain boundaries.However,the mechanism responsible for fine-grain strengthening in ferroelectric materials remains unclear,primarily due to the complex multi-field coupling effect arising from spontaneous polarization.Through molecular dynamics simulations,we investigate the strengthening mechanism of barium titanate(BaTiO3),with extremely fine-grain sizes.This material exhibits an inverse Hall–Petch relationship between grain size and strength,rooting in the inhomogeneous concentration of atomic strain and grain rotation.Furthermore,we present a theoretical model to predict the transition from the inverse Hall–Petch stage to the Hall–Petch stage based on strength variations with size,which aligns well with the simulation results.It has been found that the piezoelectric properties of the BaTiO3 are affected by polarization domain switching at various grain sizes.This study enhances our understanding of the atomic-scale mechanisms that contribute to the performance evolution of fine-grain nano-ferroelectric materials.It also provides valuable insights into the design of extremely small-scale ferroelectric components.展开更多
Fine-grained BaTiO3-based X7R ceramic materials were prepared and the effects of milling process on the core-shell structures and dielectric properties were investigated using scanning electron microscope, transmissio...Fine-grained BaTiO3-based X7R ceramic materials were prepared and the effects of milling process on the core-shell structures and dielectric properties were investigated using scanning electron microscope, transmission electron microscope, and energy dispersive spectroscopy （EDS）. As the milling time extends, the dielectric constant of the ceramics increases, whereas the temperature coefficient of capacitance at 125℃ drops quickly. The changes in dielectric properties are considered relevant to the microstructure evolution caused by the milling process. Defects on the surface of BaTiO3 particles increase because of the effects of milling process, which will make it easier for additives to diffuse into the interior grains. As the milling time increases, the shell region gets thicker and the core region gets smaller; however, EDS results show that the chemical inhomogeneity between grain core and grain shell becomes weaker.展开更多
The effect of Sb on the microstructure and mechanical properties of Mg2Si/Al-Si composites was investigated.The results show that Sb can improve the microstructure and mechanical properties of Mg2Si/Al-Si composites.W...The effect of Sb on the microstructure and mechanical properties of Mg2Si/Al-Si composites was investigated.The results show that Sb can improve the microstructure and mechanical properties of Mg2Si/Al-Si composites.When the content of Sb is 0.4%,the morphology of primary Mg2Si changes from dendrites to fine particles,the average size of Mg2Si particles is refined from 52 to 25μm,and the ultimate tensile strength and elongation of the composites increase from 102.1 MPa and 0.26% to 138.6 MPa and 0.36%,respectively.The strengthening mechanism can be attributed to the fine-grain strengthening.However,excessive Sb is disadvantageous to the modification of the composites.展开更多
Face forgery detection is drawing ever-increasing attention in the academic community owing to security concerns.Despite the considerable progress in existing methods,we note that:Previous works overlooked finegrain f...Face forgery detection is drawing ever-increasing attention in the academic community owing to security concerns.Despite the considerable progress in existing methods,we note that:Previous works overlooked finegrain forgery cues with high transferability.Such cues positively impact the model’s accuracy and generalizability.Moreover,single-modality often causes overfitting of the model,and Red-Green-Blue(RGB)modal-only is not conducive to extracting the more detailed forgery traces.We propose a novel framework for fine-grain forgery cues mining with fusion modality to cope with these issues.First,we propose two functional modules to reveal and locate the deeper forged features.Our method locates deeper forgery cues through a dual-modality progressive fusion module and a noise adaptive enhancement module,which can excavate the association between dualmodal space and channels and enhance the learning of subtle noise features.A sensitive patch branch is introduced on this foundation to enhance the mining of subtle forgery traces under fusion modality.The experimental results demonstrate that our proposed framework can desirably explore the differences between authentic and forged images with supervised learning.Comprehensive evaluations of several mainstream datasets show that our method outperforms the state-of-the-art detection methods with remarkable detection ability and generalizability.展开更多
文摘跨模态行人重识别研究的重难点主要来自于行人图像之间巨大的模态差异和模态内差异。针对这些问题,提出一种结合多尺度特征与混淆学习的网络结构。为实现高效的特征提取、缩小模态内差异,将网络设计为多尺度特征互补的形式,分别学习行人的局部细化特征与全局粗糙特征,从细粒度和粗粒度两方面来增强网络的特征表达能力。利用混淆学习策略,模糊网络的模态识别反馈,挖掘稳定且有效的模态无关属性应对模态差异,来提高特征对模态变化的鲁棒性。在大规模数据集SYSU-MM01的全搜索模式下该算法首位击中率和平均精度(mean average precision,mAP)的结果分别为76.69%和72.45%,在RegDB数据集的可见光到红外模式下该算法首位击中率和mAP的结果分别为94.62%和94.60%,优于现有的主要方法,验证了所提方法的有效性。
基金supported by the National Natural Science Foundation of China(Nos.12172117,12372154)National Science and Technology Major Project(No.J2019-1II-0010-0054)+1 种基金National Numerical Windtunnel(No.NNW2019-JT01-023)High-Performance Computing Center of Hebei University。
文摘Tailoring grain size can improve the strength of polycrystals by regulating the proportion of grains to grain boundaries and the interaction area.As the grain size decreases to the nanoscale,the deformation mechanism in polycrystals shifts from being primarily mediated by dislocations to deformation occurring within the grains and grain boundaries.However,the mechanism responsible for fine-grain strengthening in ferroelectric materials remains unclear,primarily due to the complex multi-field coupling effect arising from spontaneous polarization.Through molecular dynamics simulations,we investigate the strengthening mechanism of barium titanate(BaTiO3),with extremely fine-grain sizes.This material exhibits an inverse Hall–Petch relationship between grain size and strength,rooting in the inhomogeneous concentration of atomic strain and grain rotation.Furthermore,we present a theoretical model to predict the transition from the inverse Hall–Petch stage to the Hall–Petch stage based on strength variations with size,which aligns well with the simulation results.It has been found that the piezoelectric properties of the BaTiO3 are affected by polarization domain switching at various grain sizes.This study enhances our understanding of the atomic-scale mechanisms that contribute to the performance evolution of fine-grain nano-ferroelectric materials.It also provides valuable insights into the design of extremely small-scale ferroelectric components.
基金supported by the National Science fund for Distinguished Young Scholars (No.50625204)the National Natural Science Foundation of China (Science Fund for Creative Research Groups)(No.50621201)+1 种基金the Major State Basic Research Development Program of China (No.2009CB623301)the National High-Tech Research and Development Program of China (No.2006AA03Z0428), and Samsung Electro-Mechanics Co., Ltd.
文摘Fine-grained BaTiO3-based X7R ceramic materials were prepared and the effects of milling process on the core-shell structures and dielectric properties were investigated using scanning electron microscope, transmission electron microscope, and energy dispersive spectroscopy （EDS）. As the milling time extends, the dielectric constant of the ceramics increases, whereas the temperature coefficient of capacitance at 125℃ drops quickly. The changes in dielectric properties are considered relevant to the microstructure evolution caused by the milling process. Defects on the surface of BaTiO3 particles increase because of the effects of milling process, which will make it easier for additives to diffuse into the interior grains. As the milling time increases, the shell region gets thicker and the core region gets smaller; however, EDS results show that the chemical inhomogeneity between grain core and grain shell becomes weaker.
文摘The effect of Sb on the microstructure and mechanical properties of Mg2Si/Al-Si composites was investigated.The results show that Sb can improve the microstructure and mechanical properties of Mg2Si/Al-Si composites.When the content of Sb is 0.4%,the morphology of primary Mg2Si changes from dendrites to fine particles,the average size of Mg2Si particles is refined from 52 to 25μm,and the ultimate tensile strength and elongation of the composites increase from 102.1 MPa and 0.26% to 138.6 MPa and 0.36%,respectively.The strengthening mechanism can be attributed to the fine-grain strengthening.However,excessive Sb is disadvantageous to the modification of the composites.
基金This study is supported by the Fundamental Research Funds for the Central Universities of PPSUC under Grant 2022JKF02009.
文摘Face forgery detection is drawing ever-increasing attention in the academic community owing to security concerns.Despite the considerable progress in existing methods,we note that:Previous works overlooked finegrain forgery cues with high transferability.Such cues positively impact the model’s accuracy and generalizability.Moreover,single-modality often causes overfitting of the model,and Red-Green-Blue(RGB)modal-only is not conducive to extracting the more detailed forgery traces.We propose a novel framework for fine-grain forgery cues mining with fusion modality to cope with these issues.First,we propose two functional modules to reveal and locate the deeper forged features.Our method locates deeper forgery cues through a dual-modality progressive fusion module and a noise adaptive enhancement module,which can excavate the association between dualmodal space and channels and enhance the learning of subtle noise features.A sensitive patch branch is introduced on this foundation to enhance the mining of subtle forgery traces under fusion modality.The experimental results demonstrate that our proposed framework can desirably explore the differences between authentic and forged images with supervised learning.Comprehensive evaluations of several mainstream datasets show that our method outperforms the state-of-the-art detection methods with remarkable detection ability and generalizability.
