Speech-face association aims to achieve identity matching between facial images and voice segments by aligning cross-modal features.Existing research primarily focuses on learning shared-space representations and comp...Speech-face association aims to achieve identity matching between facial images and voice segments by aligning cross-modal features.Existing research primarily focuses on learning shared-space representations and computing one-to-one similarities between cross-modal sample pairs to establish their correlation.However,these approaches do not fully account for intra-class variations between the modalities or the many-to-many relationships among cross-modal samples,which are crucial for robust association modeling.To address these challenges,we propose a novel framework that leverages global information to align voice and face embeddings while effectively correlating identity information embedded in both modalities.First,we jointly pre-train face recognition and speaker recognition networks to encode discriminative features from facial images and voice segments.This shared pre-training step ensures the extraction of complementary identity information across modalities.Subsequently,we introduce a cross-modal simplex center loss,which aligns samples with identity centers located at the vertices of a regular simplex inscribed on a hypersphere.This design enforces an equidistant and balanced distribution of identity embeddings,reducing intra-class variations.Furthermore,we employ an improved triplet center loss that emphasizes hard sample mining and optimizes inter-class separability,enhancing the model’s ability to generalize across challenging scenarios.Extensive experiments validate the effectiveness of our framework,demonstrating superior performance across various speech-face association tasks,including matching,verification,and retrieval.Notably,in the challenging gender-constrained matching task,our method achieves a remarkable accuracy of 79.22%,significantly outperforming existing approaches.These results highlight the potential of the proposed framework to advance the state of the art in cross-modal identity association.展开更多
Exosomes have shown good potential in ischemic injury disease treatments.However,evidence about their effect and molecular mechanisms in osteonecrosis of femoral head(ONFH)treatment is still limited.Here,we revealed t...Exosomes have shown good potential in ischemic injury disease treatments.However,evidence about their effect and molecular mechanisms in osteonecrosis of femoral head(ONFH)treatment is still limited.Here,we revealed the cell biology characters of ONFH osteonecrosis area bone tissue in single cell scale and thus identified a novel ONFH treatment approach based on M2 macrophages-derived exosomes(M2-Exos).We further show that M2-Exos are highly effective in the treatment of ONFH by modulating the phenotypes communication between neutrophil and endothelium including neutrophil extracellular traps formation and endothelial phenotype transition.Additionally,we identified that M2-Exos’therapeutic effect is attributed to the high content of miR-93-5p and constructed miR-93-5p overexpression model in vitro and in vivo based on lentivirus and adenoassociated virus respectively.Then we found miR-93-5p can not only reduce neutrophil extracellular traps formation but also improve angiogenic ability of endothelial cells.These results provided a new theoretical basis for the clinical application of ONFH therapeutic exosomes.展开更多
Emotion recognition under uncontrolled and noisy environments presents persistent challenges in the design of emotionally responsive systems.The current study introduces an audio-visual recognition framework designed ...Emotion recognition under uncontrolled and noisy environments presents persistent challenges in the design of emotionally responsive systems.The current study introduces an audio-visual recognition framework designed to address performance degradation caused by environmental interference,such as background noise,overlapping speech,and visual obstructions.The proposed framework employs a structured fusion approach,combining early-stage feature-level integration with decision-level coordination guided by temporal attention mechanisms.Audio data are transformed into mel-spectrogram representations,and visual data are represented as raw frame sequences.Spatial and temporal features are extracted through convolutional and transformer-based encoders,allowing the framework to capture complementary and hierarchical information fromboth sources.Across-modal attentionmodule enables selective emphasis on relevant signals while suppressing modality-specific noise.Performance is validated on a modified version of the AFEW dataset,in which controlled noise is introduced to emulate realistic conditions.The framework achieves higher classification accuracy than comparative baselines,confirming increased robustness under conditions of cross-modal disruption.This result demonstrates the suitability of the proposed method for deployment in practical emotion-aware technologies operating outside controlled environments.The study also contributes a systematic approach to fusion design and supports further exploration in the direction of resilientmultimodal emotion analysis frameworks.The source code is publicly available at https://github.com/asmoon002/AVER(accessed on 18 August 2025).展开更多
Detecting surface defects on unused rails is crucial for evaluating rail quality and durability to ensure the safety of rail transportation.However,existing detection methods often struggle with challenges such as com...Detecting surface defects on unused rails is crucial for evaluating rail quality and durability to ensure the safety of rail transportation.However,existing detection methods often struggle with challenges such as complex defect morphology,texture similarity,and fuzzy edges,leading to poor accuracy and missed detections.In order to resolve these problems,we propose MSCM-Net(Multi-Scale Cross-Modal Network),a multiscale cross-modal framework focused on detecting rail surface defects.MSCM-Net introduces an attention mechanism to dynamically weight the fusion of RGB and depth maps,effectively capturing and enhancing features at different scales for each modality.To further enrich feature representation and improve edge detection in blurred areas,we propose a multi-scale void fusion module that integrates multi-scale feature information.To improve cross-modal feature fusion,we develop a cross-enhanced fusion module that transfers fused features between layers to incorporate interlayer information.We also introduce a multimodal feature integration module,which merges modality-specific features from separate decoders into a shared decoder,enhancing detection by leveraging richer complementary information.Finally,we validate MSCM-Net on the NEU RSDDS-AUG RGB-depth dataset,comparing it against 12 leading methods,and the results show that MSCM-Net achieves superior performance on all metrics.展开更多
With the rapid growth of socialmedia,the spread of fake news has become a growing problem,misleading the public and causing significant harm.As social media content is often composed of both images and text,the use of...With the rapid growth of socialmedia,the spread of fake news has become a growing problem,misleading the public and causing significant harm.As social media content is often composed of both images and text,the use of multimodal approaches for fake news detection has gained significant attention.To solve the problems existing in previous multi-modal fake news detection algorithms,such as insufficient feature extraction and insufficient use of semantic relations between modes,this paper proposes the MFFFND-Co(Multimodal Feature Fusion Fake News Detection with Co-Attention Block)model.First,the model deeply explores the textual content,image content,and frequency domain features.Then,it employs a Co-Attention mechanism for cross-modal fusion.Additionally,a semantic consistency detectionmodule is designed to quantify semantic deviations,thereby enhancing the performance of fake news detection.Experimentally verified on two commonly used datasets,Twitter and Weibo,the model achieved F1 scores of 90.0% and 94.0%,respectively,significantly outperforming the pre-modified MFFFND(Multimodal Feature Fusion Fake News Detection with Attention Block)model and surpassing other baseline models.This improves the accuracy of detecting fake information in artificial intelligence detection and engineering software detection.展开更多
Regulating the electronic structure and oxygencontaining intermediates adsorption behavior on Fe-based catalysts is of great significance to cope with the sluggish oxygen reduction reaction(ORR)kinetics,but it still r...Regulating the electronic structure and oxygencontaining intermediates adsorption behavior on Fe-based catalysts is of great significance to cope with the sluggish oxygen reduction reaction(ORR)kinetics,but it still remains a great challenge.In this work,Fe atom clusters(Fe_(AC))modified by high-density Cu single atoms(Cu_(SA))in a N,S-doped porous carbon substrate(Fe_(AC)/Cu_(SA)@NCS)is reported for enhanced ORR electrocatalysis.Fe_(AC)/Cu_(SA)@NCS exhibits excellent ORR performance with a half-wave potential(E_(1/2))of 0.911 V,a high four-electron process selectivity and excellent stability.The ORR performance is also verified in the Fe_(AC)/Cu_(SA)@NCS-based Zn-air battery,which shows a high peak power density of 192.67 mW cm^(-2),a higher specific capacity of 808.3 mAh g^(-1)and impressive charge-discharge cycle stability.Moreover,density functional theory calculations show that Cu single atoms synergistically modulate the electronic structure Fe active atoms in Fe atomic clusters,reducing the energy barrier of the rate-determining step(i.e.,*OH desorption)on Fe_(AC)/Cu_(SA)@NCS.This work provides an effective way to regulate the electronic structure of Fe-based catalysts and optimize their electrocatalytic activity based on the introduction of a second metal source.展开更多
While reinforcement learning-based underwater acoustic adaptive modulation shows promise for enabling environment-adaptive communication as supported by extensive simulation-based research,its practical performance re...While reinforcement learning-based underwater acoustic adaptive modulation shows promise for enabling environment-adaptive communication as supported by extensive simulation-based research,its practical performance remains underexplored in field investigations.To evaluate the practical applicability of this emerging technique in adverse shallow sea channels,a field experiment was conducted using three communication modes:orthogonal frequency division multiplexing(OFDM),M-ary frequency-shift keying(MFSK),and direct sequence spread spectrum(DSSS)for reinforcement learning-driven adaptive modulation.Specifically,a Q-learning method is used to select the optimal modulation mode according to the channel quality quantified by signal-to-noise ratio,multipath spread length,and Doppler frequency offset.Experimental results demonstrate that the reinforcement learning-based adaptive modulation scheme outperformed fixed threshold detection in terms of total throughput and average bit error rate,surpassing conventional adaptive modulation strategies.展开更多
Realizing the valley Hall effect by breaking the spatial inversion symmetry of photonic systems has become a cutting-edge field of micro-nano-optics,since the valley degree of freedom was introduced into photonic syst...Realizing the valley Hall effect by breaking the spatial inversion symmetry of photonic systems has become a cutting-edge field of micro-nano-optics,since the valley degree of freedom was introduced into photonic system.Various novel devices based on the domain walls of the valley photonic crystals have also been demonstrated.In this article,we investigate the variation of edge states by the modulation of refractive index within the domain walls,and the geometric difference between the dielectric columns of the sublattices.Straight photonic crystal waveguides with three types of domain walls(bearded,zigzag,armchair)are constructed.Simulation results show that the creation of a double-edge state in the band diagram results in two windows of stable transmission in tunable bands.Our findings might have significant implications in the field of novel optical devices.展开更多
Single-atom catalysts are promising for H_(2)O_(2) photosynthesis from O_(2) and H_(2)O,but their efficiency is still limited by the ill-defined electronic structure.In this study,Co single-atoms with unique four plan...Single-atom catalysts are promising for H_(2)O_(2) photosynthesis from O_(2) and H_(2)O,but their efficiency is still limited by the ill-defined electronic structure.In this study,Co single-atoms with unique four planar N-coordination and one axial P-coordination(Co-N_(4)P_(1))are decorated on the lateral edges of nanorod-like crystalline g-C_(3)N_(4)(CCN)photocatalysts.Significantly,the electronic structures of central Co as active sites for O_(2) reduction reaction(ORR)and planar N-coordinator as active sites for H_(2)O oxidation reaction(WOR)in Co-N_(4)P_(1) can be well regulated by the synergetic effects of introducing axial P-coordinator,in contrast to the decorated Co single-atoms with only four planar N-coordination(Co-N_(4)).Specifically,directional photoelectron accumulation at central Co active sites,induced by an introduced midgap level in Co-N_(4)P_(1),mediates the ORR active sites from 4e–-ORR-selective terminal–NH_(2) sites to 2e–-ORR-selective Co sites,moreover,an elevated d-band center of Co 3d orbital strengthens ORR intermediate*OOH adsorption,thus jointly facilitating a highly selective and active 2e^(–)-ORR pathway to H_(2)O_(2) photosynthesis.Simultaneously,a downshifted p-band center of N_(2)p orbital in Co-N_(4)P_(1) weakens WOR intermediate*OH adsorption,thus enabling a preferable 2e^(–)-WOR pathway toward H_(2)O_(2) photosynthesis.Subsequently,Co-N_(4)P_(1) exhibits exceptional H_(2)O_(2) photosynthesis efficiency,reaching 295.6μmol g^(-1) h^(-1) with a remarkable solar-to-chemical conversion efficiency of 0.32%,which is 15 times that of Co-N_(4)(19.2μmol g^(-1) h^(-1))and 10 times higher than CCN(27.6μmol g^(-1) h^(-1)).This electronic structure modulation on single-atom catalysts offers a promising strategy for boosting the activity and selectivity of H_(2)O_(2) photosynthesis.展开更多
Lysophosphatidic acid(LPA)is a pleiotropic lipid agonist essential for functions of the central nervous system(CNS).It is abundant in the developing and adult brain while its concentration in biological fluids,includi...Lysophosphatidic acid(LPA)is a pleiotropic lipid agonist essential for functions of the central nervous system(CNS).It is abundant in the developing and adult brain while its concentration in biological fluids,including cerebrospinal fluid,varies significantly(Figure 1Α;Yung et al.,2014).LPA actually corresponds to a variety of lipid species that include different stereoisomers with either saturated or unsaturated fatty acids bearing likely differentiated biological activities(Figure 1Α;Yung et al.,2014;Hernández-Araiza et al.,2018).展开更多
The catalytic performance of solid catalysts depends on the properties of the catalytically active sites and their accessibility to reactants, which are significantly affected by the microstructure（morphology, shape,...The catalytic performance of solid catalysts depends on the properties of the catalytically active sites and their accessibility to reactants, which are significantly affected by the microstructure（morphology, shape, size, texture, and surface structure） and surface chemistry（elemental components and chemical states）. The development of facile and efficient methods for tailoring the microstructure and surface chemistry is a hot topic in catalysis. This contribution reviews the state of the art in modulating the microstructure and surface chemistry of carbocatalysts by both bottom‐up and top‐down strategies and their use in the oxidative dehydrogenation（ODH） and direct dehydrogenation（DDH） of hydrocarbons including light alkanes and ethylbenzene to their corresponding olefins, important building blocks and chemicals like oxygenates. A concept of microstructure and surface chemistry tuning of the carbocatalyst for optimized catalytic performance and also for the fundamental understanding of the structure‐performance relationship is discussed. We also highlight the importance and challenges in modulating the microstructure and surface chemistry of carbocatalysts in ODH and DDH reactions of hydrocarbons for the highly‐efficient, energy‐saving,and clean production of their corresponding olefins.展开更多
An approach to identification of linear continuous-time system is studied with modulating functions. Based on wavelet analysis theory, the multi-resolution modulating functions are designed, and the corresponding filt...An approach to identification of linear continuous-time system is studied with modulating functions. Based on wavelet analysis theory, the multi-resolution modulating functions are designed, and the corresponding filters have been analyzed. Using linear modulating filters, we can obtain an identification model that is parameterized directly in continuous-time model parameters. By applying the results from discrete-time model identification to the obtained identification model, a continuous-time estimation method is developed. Considering the accuracy of parameter estimates, an instrumental variable (Ⅳ) method is proposed, and the design of modulating integral filter is discussed. The relationship between the accuracy of identification and the parameter of modulating filter is investigated, and some points about designing Gaussian wavelet modulating function are outlined. Finally, a simulation study is also included to verify the theoretical results.展开更多
Traumatic injury to the adult mammalian central nervous system（CNS） leads to complex cellular responses. Among them, the scar tissue formed is generally recognized as a major obstacle to CNS repair, both by the prod...Traumatic injury to the adult mammalian central nervous system（CNS） leads to complex cellular responses. Among them, the scar tissue formed is generally recognized as a major obstacle to CNS repair, both by the production of inhibitory molecules and by the physical impedance of axon regrowth. Therefore, scar-modulating treatments have become a leading therapeutic intervention for CNS injury. To date, a variety of biological and pharmaceutical treatments, targeting scar modulation, have been tested in animal models of CNS injury, and a few are likely to enter clinical trials. In this review, we summarize current knowledge of the scar-modulating treatments according to their specific aims：（1） inhibition of glial and fibrotic scar formation, and（2） blockade of the production of scar-associated inhibitory molecules. The removal of existing scar tissue is also discussed as a treatment of choice. It is believed that only a combinatorial strategy is likely to help eliminate the detrimental effects of scar tissue on CNS repair.展开更多
Improving peroral delivery efficiency is always a persistent goal for both small-molecule and macromolecular drug development. However,intestinal mucus barrier which greatly impedes drug-loaded nanoparticles penetrati...Improving peroral delivery efficiency is always a persistent goal for both small-molecule and macromolecular drug development. However,intestinal mucus barrier which greatly impedes drug-loaded nanoparticles penetration is commonly overlooked. Therefore,in this study,taking fluorescent labeled PLGA(poly(lactic-co-glycolic acid)) nanoparticles as a tool,the influence of anionic and nonionic surfactants on mucus penetration ability of nanoparticles and their mucus barrier regulating ability were studied. The movement of PLGA nanoparticles in mucus was tracked by multiple particles tracking method(MPT).Alteration of mucus properties by addition of surfactants was evaluated by rheology and morphology study. Rat intestinal villus penetration study was used to further evaluate penetration enhancement of nanoparticles. The effective diffusivities of the nanoparticles in surfactants pretreated mucus were increased by 2–3 times and the mucus barrier regulating capacity was also surfactant type dependent. Sodium dodecyl sulfate(SDS) increased the complex viscosity and viscoelastic properties of mucus,but poloxamer presented a decreased trend. Tween 80 maintained the rheological property of the mucus. With the mucus barrier regulated by surfactants,the penetration of nanoparticles in intestinal villus was obviously increased. In summary,the mucus penetration ability of nanoparticles could be enhanced by altering mucus microenvironment with surfactants. Tween 80 which largely retains the original mucus rheology and morphology properties may be a promising candidate for facilitating nanoparticle penetration through the mucus barrier with good safety profile.展开更多
This work experimentally demonstrates a new method of optimizing the transport of cold atoms via modulating the velocity profile imposed on a magnetic quadrupole trap.The trap velocity and corresponding modulation are...This work experimentally demonstrates a new method of optimizing the transport of cold atoms via modulating the velocity profile imposed on a magnetic quadrupole trap.The trap velocity and corresponding modulation are controlled by varying the currents of two pairs of anti-Helmholtz coils.Cold 87Rb atoms are transported in a non-adiabatic regime over 22 mm in 200 ms.For the transported atoms their final-vibration amplitude dependences of modulation period number,depth,and initial phase are investigated.With modulation period n = 5,modulation depth K = 0.55,and initial phase φ = 0,cold atom clouds with more atom numbers,smaller final-vibration amplitude,and lower temperature are efficiently transported.Theoretical analysis and numerical simulation are also provided,which are in good agreement with experimental results.展开更多
The biggest challenge is to develop a low cost and readily available catalyst to replace expensive commercial Pt/C for efficient electrochemical oxygen reduction reaction(ORR).In this research,closo-[B_(12)H_(12)]^(2−...The biggest challenge is to develop a low cost and readily available catalyst to replace expensive commercial Pt/C for efficient electrochemical oxygen reduction reaction(ORR).In this research,closo-[B_(12)H_(12)]^(2−)and 1,10-phenanthroline-iron complexes were introduced into the porous metal-organic framework by impregnation method,and further annealing treatment achieved the successful anchoring of single-atom-Fe in B-doped CN Matrix(FeN4CB).The ORR activity of FeN4CB is comparable to the widely used commercial 20 wt%Pt/C.Where the half-wave potential(E_(1/2))in alkaline medium up to 0.84 V,and even in the face of challenging ORR in acidic medium,the E_(1/2)of ORR driven by FeN4CB is still as high as 0.81 V.When FeN4CB was used as air cathode,the open circuit voltage of Zn-air battery reaches 1.435 V,and the power density and specific capacity are as high as 177 mW cm^(−2)and 800 mAh g_(Zn)^(−1)(theoretical value:820 mAh g_(Zn)^(−1)),respectively.The dazzling point of FeN4CB also appears in the high ORR stability,whether in alkaline or acidic media,E_(1/2)and limiting current density are still close to the initial value after 5000 times cycles.After continuously running the charge-discharge test for 220 h,the charge voltage and discharge voltage of the rechargeable zinc-air battery with FeN4CB as the air cathode maintained the initial state.Density functional theory calculations reveals that introducing B atom to Fe–N4–C can adjust the electronic structure to easily break O=O bond and significantly reduce the energy barrier of the rate-determining step resulting in an improved ORR activity.展开更多
The neuromodulatory transmitter serotonin(5-hydroxytryptamine,5-HT)is synthesized by neurons located in the brainstem,which project more or less densely to the entire central nervous system(Charnay and Leger,2010).Ser...The neuromodulatory transmitter serotonin(5-hydroxytryptamine,5-HT)is synthesized by neurons located in the brainstem,which project more or less densely to the entire central nervous system(Charnay and Leger,2010).Serotonin regulates a variety of physiological functions,including food intake,reward,reproduction,sleep-wake cycle,memory,cognition,emotion,and mood(Charnay and Leger,2010).展开更多
The electronic structure of catalytic active sites can be influenced by modulating the coordination bonding of the central single metal atom,but it is difficult to achieve.Herein,we reported the single Zn-atom incorpo...The electronic structure of catalytic active sites can be influenced by modulating the coordination bonding of the central single metal atom,but it is difficult to achieve.Herein,we reported the single Zn-atom incorporated dual doped P,N carbon framework(Zn-N_(4)P/C)for ORR via engineering the surrounding coordination environment of active centers.The Zn-N_(4)P/C catalyst exhibited comparable ORR activity(E_(1/2)=0.86 V)and significantly better ORR stability than that of Pt/C catalyst.It also shows respectable performance in terms of maximum peak power density(249.6 mW cm^(-2)),specific capacitance(779 mAh g^(-1)),and charge-discharge cycling stability for 150 hours in Zn-air battery.The high catalytic activity is attributed to the uniform active sites,tunable electronic/geometric configuration,optimized intrinsic activity,and faster mass transfer during ORR-pathway.Further,theoretical results exposed that the Zn-N_(4)P configuration is more electrochemically active as compared to Zn-N_(4) structure for the oxygen reduction reaction.展开更多
Multimodal sentiment analysis aims to understand people’s emotions and opinions from diverse data.Concate-nating or multiplying various modalities is a traditional multi-modal sentiment analysis fusion method.This fu...Multimodal sentiment analysis aims to understand people’s emotions and opinions from diverse data.Concate-nating or multiplying various modalities is a traditional multi-modal sentiment analysis fusion method.This fusion method does not utilize the correlation information between modalities.To solve this problem,this paper proposes amodel based on amulti-head attention mechanism.First,after preprocessing the original data.Then,the feature representation is converted into a sequence of word vectors and positional encoding is introduced to better understand the semantic and sequential information in the input sequence.Next,the input coding sequence is fed into the transformer model for further processing and learning.At the transformer layer,a cross-modal attention consisting of a pair of multi-head attention modules is employed to reflect the correlation between modalities.Finally,the processed results are input into the feedforward neural network to obtain the emotional output through the classification layer.Through the above processing flow,the model can capture semantic information and contextual relationships and achieve good results in various natural language processing tasks.Our model was tested on the CMU Multimodal Opinion Sentiment and Emotion Intensity(CMU-MOSEI)and Multimodal EmotionLines Dataset(MELD),achieving an accuracy of 82.04% and F1 parameters reached 80.59% on the former dataset.展开更多
Studies on the integration of cross-modal information with taste perception has been mostly limited to uni-modal level.The cross-modal sensory interaction and the neural network of information processing and its contr...Studies on the integration of cross-modal information with taste perception has been mostly limited to uni-modal level.The cross-modal sensory interaction and the neural network of information processing and its control were not fully explored and the mechanisms remain poorly understood.This mini review investigated the impact of uni-modal and multi-modal information on the taste perception,from the perspective of cognitive status,such as emotion,expectation and attention,and discussed the hypothesis that the cognitive status is the key step for visual sense to exert influence on taste.This work may help researchers better understand the mechanism of cross-modal information processing and further develop neutrally-based artificial intelligent(AI)system.展开更多
基金funded by the Scientific Funding for China Academy of Railway Sciences Corporation Limited,China(No.2023YJ125).
文摘Speech-face association aims to achieve identity matching between facial images and voice segments by aligning cross-modal features.Existing research primarily focuses on learning shared-space representations and computing one-to-one similarities between cross-modal sample pairs to establish their correlation.However,these approaches do not fully account for intra-class variations between the modalities or the many-to-many relationships among cross-modal samples,which are crucial for robust association modeling.To address these challenges,we propose a novel framework that leverages global information to align voice and face embeddings while effectively correlating identity information embedded in both modalities.First,we jointly pre-train face recognition and speaker recognition networks to encode discriminative features from facial images and voice segments.This shared pre-training step ensures the extraction of complementary identity information across modalities.Subsequently,we introduce a cross-modal simplex center loss,which aligns samples with identity centers located at the vertices of a regular simplex inscribed on a hypersphere.This design enforces an equidistant and balanced distribution of identity embeddings,reducing intra-class variations.Furthermore,we employ an improved triplet center loss that emphasizes hard sample mining and optimizes inter-class separability,enhancing the model’s ability to generalize across challenging scenarios.Extensive experiments validate the effectiveness of our framework,demonstrating superior performance across various speech-face association tasks,including matching,verification,and retrieval.Notably,in the challenging gender-constrained matching task,our method achieves a remarkable accuracy of 79.22%,significantly outperforming existing approaches.These results highlight the potential of the proposed framework to advance the state of the art in cross-modal identity association.
基金the support of the National Natural Science Foundation of China (Grant No.82272503)Natural Science Foundation of Zhejiang Province (Grant No. LQN25H060006)
文摘Exosomes have shown good potential in ischemic injury disease treatments.However,evidence about their effect and molecular mechanisms in osteonecrosis of femoral head(ONFH)treatment is still limited.Here,we revealed the cell biology characters of ONFH osteonecrosis area bone tissue in single cell scale and thus identified a novel ONFH treatment approach based on M2 macrophages-derived exosomes(M2-Exos).We further show that M2-Exos are highly effective in the treatment of ONFH by modulating the phenotypes communication between neutrophil and endothelium including neutrophil extracellular traps formation and endothelial phenotype transition.Additionally,we identified that M2-Exos’therapeutic effect is attributed to the high content of miR-93-5p and constructed miR-93-5p overexpression model in vitro and in vivo based on lentivirus and adenoassociated virus respectively.Then we found miR-93-5p can not only reduce neutrophil extracellular traps formation but also improve angiogenic ability of endothelial cells.These results provided a new theoretical basis for the clinical application of ONFH therapeutic exosomes.
基金funded by the Institute of Information&CommunicationsTechnology Planning&Evaluation(IITP)grant funded by the Korea government(MSIT),grant number 2021-0-01341.
文摘Emotion recognition under uncontrolled and noisy environments presents persistent challenges in the design of emotionally responsive systems.The current study introduces an audio-visual recognition framework designed to address performance degradation caused by environmental interference,such as background noise,overlapping speech,and visual obstructions.The proposed framework employs a structured fusion approach,combining early-stage feature-level integration with decision-level coordination guided by temporal attention mechanisms.Audio data are transformed into mel-spectrogram representations,and visual data are represented as raw frame sequences.Spatial and temporal features are extracted through convolutional and transformer-based encoders,allowing the framework to capture complementary and hierarchical information fromboth sources.Across-modal attentionmodule enables selective emphasis on relevant signals while suppressing modality-specific noise.Performance is validated on a modified version of the AFEW dataset,in which controlled noise is introduced to emulate realistic conditions.The framework achieves higher classification accuracy than comparative baselines,confirming increased robustness under conditions of cross-modal disruption.This result demonstrates the suitability of the proposed method for deployment in practical emotion-aware technologies operating outside controlled environments.The study also contributes a systematic approach to fusion design and supports further exploration in the direction of resilientmultimodal emotion analysis frameworks.The source code is publicly available at https://github.com/asmoon002/AVER(accessed on 18 August 2025).
基金funded by the National Natural Science Foundation of China(grant number 62306186)the Technology Plan Joint Foundation of Liaoning Province(grant number 2023-MSLH-246)the Technology Plan Joint Foundation of Liaoning Province(grant number 2023-BSBA-238).
文摘Detecting surface defects on unused rails is crucial for evaluating rail quality and durability to ensure the safety of rail transportation.However,existing detection methods often struggle with challenges such as complex defect morphology,texture similarity,and fuzzy edges,leading to poor accuracy and missed detections.In order to resolve these problems,we propose MSCM-Net(Multi-Scale Cross-Modal Network),a multiscale cross-modal framework focused on detecting rail surface defects.MSCM-Net introduces an attention mechanism to dynamically weight the fusion of RGB and depth maps,effectively capturing and enhancing features at different scales for each modality.To further enrich feature representation and improve edge detection in blurred areas,we propose a multi-scale void fusion module that integrates multi-scale feature information.To improve cross-modal feature fusion,we develop a cross-enhanced fusion module that transfers fused features between layers to incorporate interlayer information.We also introduce a multimodal feature integration module,which merges modality-specific features from separate decoders into a shared decoder,enhancing detection by leveraging richer complementary information.Finally,we validate MSCM-Net on the NEU RSDDS-AUG RGB-depth dataset,comparing it against 12 leading methods,and the results show that MSCM-Net achieves superior performance on all metrics.
基金supported by Communication University of China(HG23035)partly supported by the Fundamental Research Funds for the Central Universities(CUC230A013).
文摘With the rapid growth of socialmedia,the spread of fake news has become a growing problem,misleading the public and causing significant harm.As social media content is often composed of both images and text,the use of multimodal approaches for fake news detection has gained significant attention.To solve the problems existing in previous multi-modal fake news detection algorithms,such as insufficient feature extraction and insufficient use of semantic relations between modes,this paper proposes the MFFFND-Co(Multimodal Feature Fusion Fake News Detection with Co-Attention Block)model.First,the model deeply explores the textual content,image content,and frequency domain features.Then,it employs a Co-Attention mechanism for cross-modal fusion.Additionally,a semantic consistency detectionmodule is designed to quantify semantic deviations,thereby enhancing the performance of fake news detection.Experimentally verified on two commonly used datasets,Twitter and Weibo,the model achieved F1 scores of 90.0% and 94.0%,respectively,significantly outperforming the pre-modified MFFFND(Multimodal Feature Fusion Fake News Detection with Attention Block)model and surpassing other baseline models.This improves the accuracy of detecting fake information in artificial intelligence detection and engineering software detection.
基金financially supported by the National Natural Science Foundation of China(No.22278042)the National Natural Science Foundation of Jiangsu Province(No.BK20240567)+2 种基金the Introduction and Cultivation of Leading Innovative Talents Foundation of Changzhou,Jiangsu Province(No.CQ20220093)the Natural Science Foundation of the Jiangsu Higher Education Institutions of China(No.24KJD530001)the Open Project Program of Key Laboratory of Optic-electric Sensing and Analytical Chemistry for Life Science(No.M2024-7),MOE
文摘Regulating the electronic structure and oxygencontaining intermediates adsorption behavior on Fe-based catalysts is of great significance to cope with the sluggish oxygen reduction reaction(ORR)kinetics,but it still remains a great challenge.In this work,Fe atom clusters(Fe_(AC))modified by high-density Cu single atoms(Cu_(SA))in a N,S-doped porous carbon substrate(Fe_(AC)/Cu_(SA)@NCS)is reported for enhanced ORR electrocatalysis.Fe_(AC)/Cu_(SA)@NCS exhibits excellent ORR performance with a half-wave potential(E_(1/2))of 0.911 V,a high four-electron process selectivity and excellent stability.The ORR performance is also verified in the Fe_(AC)/Cu_(SA)@NCS-based Zn-air battery,which shows a high peak power density of 192.67 mW cm^(-2),a higher specific capacity of 808.3 mAh g^(-1)and impressive charge-discharge cycle stability.Moreover,density functional theory calculations show that Cu single atoms synergistically modulate the electronic structure Fe active atoms in Fe atomic clusters,reducing the energy barrier of the rate-determining step(i.e.,*OH desorption)on Fe_(AC)/Cu_(SA)@NCS.This work provides an effective way to regulate the electronic structure of Fe-based catalysts and optimize their electrocatalytic activity based on the introduction of a second metal source.
基金funding from the National Key Research and Development Program of China(No.2018YFE0110000)the National Natural Science Foundation of China(No.11274259,No.11574258)the Science and Technology Commission Foundation of Shanghai(21DZ1205500)in support of the present research.
文摘While reinforcement learning-based underwater acoustic adaptive modulation shows promise for enabling environment-adaptive communication as supported by extensive simulation-based research,its practical performance remains underexplored in field investigations.To evaluate the practical applicability of this emerging technique in adverse shallow sea channels,a field experiment was conducted using three communication modes:orthogonal frequency division multiplexing(OFDM),M-ary frequency-shift keying(MFSK),and direct sequence spread spectrum(DSSS)for reinforcement learning-driven adaptive modulation.Specifically,a Q-learning method is used to select the optimal modulation mode according to the channel quality quantified by signal-to-noise ratio,multipath spread length,and Doppler frequency offset.Experimental results demonstrate that the reinforcement learning-based adaptive modulation scheme outperformed fixed threshold detection in terms of total throughput and average bit error rate,surpassing conventional adaptive modulation strategies.
基金supported by the Self-Deployment Project Research Program of the Haixi Institutes,Chinese Academy of Sciences(No.CXZX-2022-GH09)the National Natural Science Foundation of China(No.11774103)the Fujian Science&Technology Innovation Laboratory for Optoelectronic Information of China(No.2021ZR114)。
文摘Realizing the valley Hall effect by breaking the spatial inversion symmetry of photonic systems has become a cutting-edge field of micro-nano-optics,since the valley degree of freedom was introduced into photonic system.Various novel devices based on the domain walls of the valley photonic crystals have also been demonstrated.In this article,we investigate the variation of edge states by the modulation of refractive index within the domain walls,and the geometric difference between the dielectric columns of the sublattices.Straight photonic crystal waveguides with three types of domain walls(bearded,zigzag,armchair)are constructed.Simulation results show that the creation of a double-edge state in the band diagram results in two windows of stable transmission in tunable bands.Our findings might have significant implications in the field of novel optical devices.
文摘Single-atom catalysts are promising for H_(2)O_(2) photosynthesis from O_(2) and H_(2)O,but their efficiency is still limited by the ill-defined electronic structure.In this study,Co single-atoms with unique four planar N-coordination and one axial P-coordination(Co-N_(4)P_(1))are decorated on the lateral edges of nanorod-like crystalline g-C_(3)N_(4)(CCN)photocatalysts.Significantly,the electronic structures of central Co as active sites for O_(2) reduction reaction(ORR)and planar N-coordinator as active sites for H_(2)O oxidation reaction(WOR)in Co-N_(4)P_(1) can be well regulated by the synergetic effects of introducing axial P-coordinator,in contrast to the decorated Co single-atoms with only four planar N-coordination(Co-N_(4)).Specifically,directional photoelectron accumulation at central Co active sites,induced by an introduced midgap level in Co-N_(4)P_(1),mediates the ORR active sites from 4e–-ORR-selective terminal–NH_(2) sites to 2e–-ORR-selective Co sites,moreover,an elevated d-band center of Co 3d orbital strengthens ORR intermediate*OOH adsorption,thus jointly facilitating a highly selective and active 2e^(–)-ORR pathway to H_(2)O_(2) photosynthesis.Simultaneously,a downshifted p-band center of N_(2)p orbital in Co-N_(4)P_(1) weakens WOR intermediate*OH adsorption,thus enabling a preferable 2e^(–)-WOR pathway toward H_(2)O_(2) photosynthesis.Subsequently,Co-N_(4)P_(1) exhibits exceptional H_(2)O_(2) photosynthesis efficiency,reaching 295.6μmol g^(-1) h^(-1) with a remarkable solar-to-chemical conversion efficiency of 0.32%,which is 15 times that of Co-N_(4)(19.2μmol g^(-1) h^(-1))and 10 times higher than CCN(27.6μmol g^(-1) h^(-1)).This electronic structure modulation on single-atom catalysts offers a promising strategy for boosting the activity and selectivity of H_(2)O_(2) photosynthesis.
基金supported by the Hellenic Foundation for Research and Innovation,HFRI,“2nd Call for HFRI Research Projects to support Faculty Members&Researchers”Project 02667 to GL.
文摘Lysophosphatidic acid(LPA)is a pleiotropic lipid agonist essential for functions of the central nervous system(CNS).It is abundant in the developing and adult brain while its concentration in biological fluids,including cerebrospinal fluid,varies significantly(Figure 1Α;Yung et al.,2014).LPA actually corresponds to a variety of lipid species that include different stereoisomers with either saturated or unsaturated fatty acids bearing likely differentiated biological activities(Figure 1Α;Yung et al.,2014;Hernández-Araiza et al.,2018).
基金supported by the National Natural Science Foundation of China(21276041)the Program for New Century Excellent Talents in University of Ministry of Education of China(NCET-12-0079)+1 种基金the Natural Science Foundation of Liaoning Province(2015020200)the Fundamental Research Funds for the Central Universities(DUT15LK41)~~
文摘The catalytic performance of solid catalysts depends on the properties of the catalytically active sites and their accessibility to reactants, which are significantly affected by the microstructure（morphology, shape, size, texture, and surface structure） and surface chemistry（elemental components and chemical states）. The development of facile and efficient methods for tailoring the microstructure and surface chemistry is a hot topic in catalysis. This contribution reviews the state of the art in modulating the microstructure and surface chemistry of carbocatalysts by both bottom‐up and top‐down strategies and their use in the oxidative dehydrogenation（ODH） and direct dehydrogenation（DDH） of hydrocarbons including light alkanes and ethylbenzene to their corresponding olefins, important building blocks and chemicals like oxygenates. A concept of microstructure and surface chemistry tuning of the carbocatalyst for optimized catalytic performance and also for the fundamental understanding of the structure‐performance relationship is discussed. We also highlight the importance and challenges in modulating the microstructure and surface chemistry of carbocatalysts in ODH and DDH reactions of hydrocarbons for the highly‐efficient, energy‐saving,and clean production of their corresponding olefins.
基金This project was supported by China Postdoctoral Science Foundation (2003034466)Scientific Research Fund of Hunan Provincial Education Department (02B032).
文摘An approach to identification of linear continuous-time system is studied with modulating functions. Based on wavelet analysis theory, the multi-resolution modulating functions are designed, and the corresponding filters have been analyzed. Using linear modulating filters, we can obtain an identification model that is parameterized directly in continuous-time model parameters. By applying the results from discrete-time model identification to the obtained identification model, a continuous-time estimation method is developed. Considering the accuracy of parameter estimates, an instrumental variable (Ⅳ) method is proposed, and the design of modulating integral filter is discussed. The relationship between the accuracy of identification and the parameter of modulating filter is investigated, and some points about designing Gaussian wavelet modulating function are outlined. Finally, a simulation study is also included to verify the theoretical results.
基金supported by grants from the National Basic Research Development Program (973 Program) of China (2014CB542202)National High-Technology Research Program (863 Program) of China (2012AA020502)+1 种基金the National Natural Science Foundation of China (81130080)a Project Funded by the Priority Academic Program Development of Jiangsu Higher Education Institutions,Jiangsu Province,China
文摘Traumatic injury to the adult mammalian central nervous system（CNS） leads to complex cellular responses. Among them, the scar tissue formed is generally recognized as a major obstacle to CNS repair, both by the production of inhibitory molecules and by the physical impedance of axon regrowth. Therefore, scar-modulating treatments have become a leading therapeutic intervention for CNS injury. To date, a variety of biological and pharmaceutical treatments, targeting scar modulation, have been tested in animal models of CNS injury, and a few are likely to enter clinical trials. In this review, we summarize current knowledge of the scar-modulating treatments according to their specific aims：（1） inhibition of glial and fibrotic scar formation, and（2） blockade of the production of scar-associated inhibitory molecules. The removal of existing scar tissue is also discussed as a treatment of choice. It is believed that only a combinatorial strategy is likely to help eliminate the detrimental effects of scar tissue on CNS repair.
基金financially supported by the National Natural Science Foundation of China (Grant No. 31870987)
文摘Improving peroral delivery efficiency is always a persistent goal for both small-molecule and macromolecular drug development. However,intestinal mucus barrier which greatly impedes drug-loaded nanoparticles penetration is commonly overlooked. Therefore,in this study,taking fluorescent labeled PLGA(poly(lactic-co-glycolic acid)) nanoparticles as a tool,the influence of anionic and nonionic surfactants on mucus penetration ability of nanoparticles and their mucus barrier regulating ability were studied. The movement of PLGA nanoparticles in mucus was tracked by multiple particles tracking method(MPT).Alteration of mucus properties by addition of surfactants was evaluated by rheology and morphology study. Rat intestinal villus penetration study was used to further evaluate penetration enhancement of nanoparticles. The effective diffusivities of the nanoparticles in surfactants pretreated mucus were increased by 2–3 times and the mucus barrier regulating capacity was also surfactant type dependent. Sodium dodecyl sulfate(SDS) increased the complex viscosity and viscoelastic properties of mucus,but poloxamer presented a decreased trend. Tween 80 maintained the rheological property of the mucus. With the mucus barrier regulated by surfactants,the penetration of nanoparticles in intestinal villus was obviously increased. In summary,the mucus penetration ability of nanoparticles could be enhanced by altering mucus microenvironment with surfactants. Tween 80 which largely retains the original mucus rheology and morphology properties may be a promising candidate for facilitating nanoparticle penetration through the mucus barrier with good safety profile.
基金Project supported by the National Natural Science Foundation of China (Grant No. 10974210)the National Basic Research Program of China (Grant No. 2011CB921504)
文摘This work experimentally demonstrates a new method of optimizing the transport of cold atoms via modulating the velocity profile imposed on a magnetic quadrupole trap.The trap velocity and corresponding modulation are controlled by varying the currents of two pairs of anti-Helmholtz coils.Cold 87Rb atoms are transported in a non-adiabatic regime over 22 mm in 200 ms.For the transported atoms their final-vibration amplitude dependences of modulation period number,depth,and initial phase are investigated.With modulation period n = 5,modulation depth K = 0.55,and initial phase φ = 0,cold atom clouds with more atom numbers,smaller final-vibration amplitude,and lower temperature are efficiently transported.Theoretical analysis and numerical simulation are also provided,which are in good agreement with experimental results.
基金financially supported by the NSFC-Yunnan Joint Foundation(U2002213)the Double Tops Joint Fund of the Yunnan Science and Technology Bureau and Yunnan University(2019FY003025)the‘Double-First Class’University Construction Project(C176220100042 and CZ21623201)。
文摘The biggest challenge is to develop a low cost and readily available catalyst to replace expensive commercial Pt/C for efficient electrochemical oxygen reduction reaction(ORR).In this research,closo-[B_(12)H_(12)]^(2−)and 1,10-phenanthroline-iron complexes were introduced into the porous metal-organic framework by impregnation method,and further annealing treatment achieved the successful anchoring of single-atom-Fe in B-doped CN Matrix(FeN4CB).The ORR activity of FeN4CB is comparable to the widely used commercial 20 wt%Pt/C.Where the half-wave potential(E_(1/2))in alkaline medium up to 0.84 V,and even in the face of challenging ORR in acidic medium,the E_(1/2)of ORR driven by FeN4CB is still as high as 0.81 V.When FeN4CB was used as air cathode,the open circuit voltage of Zn-air battery reaches 1.435 V,and the power density and specific capacity are as high as 177 mW cm^(−2)and 800 mAh g_(Zn)^(−1)(theoretical value:820 mAh g_(Zn)^(−1)),respectively.The dazzling point of FeN4CB also appears in the high ORR stability,whether in alkaline or acidic media,E_(1/2)and limiting current density are still close to the initial value after 5000 times cycles.After continuously running the charge-discharge test for 220 h,the charge voltage and discharge voltage of the rechargeable zinc-air battery with FeN4CB as the air cathode maintained the initial state.Density functional theory calculations reveals that introducing B atom to Fe–N4–C can adjust the electronic structure to easily break O=O bond and significantly reduce the energy barrier of the rate-determining step resulting in an improved ORR activity.
基金the German Research Foundation(DFG)grant GA 654/14-1 to OG.
文摘The neuromodulatory transmitter serotonin(5-hydroxytryptamine,5-HT)is synthesized by neurons located in the brainstem,which project more or less densely to the entire central nervous system(Charnay and Leger,2010).Serotonin regulates a variety of physiological functions,including food intake,reward,reproduction,sleep-wake cycle,memory,cognition,emotion,and mood(Charnay and Leger,2010).
文摘The electronic structure of catalytic active sites can be influenced by modulating the coordination bonding of the central single metal atom,but it is difficult to achieve.Herein,we reported the single Zn-atom incorporated dual doped P,N carbon framework(Zn-N_(4)P/C)for ORR via engineering the surrounding coordination environment of active centers.The Zn-N_(4)P/C catalyst exhibited comparable ORR activity(E_(1/2)=0.86 V)and significantly better ORR stability than that of Pt/C catalyst.It also shows respectable performance in terms of maximum peak power density(249.6 mW cm^(-2)),specific capacitance(779 mAh g^(-1)),and charge-discharge cycling stability for 150 hours in Zn-air battery.The high catalytic activity is attributed to the uniform active sites,tunable electronic/geometric configuration,optimized intrinsic activity,and faster mass transfer during ORR-pathway.Further,theoretical results exposed that the Zn-N_(4)P configuration is more electrochemically active as compared to Zn-N_(4) structure for the oxygen reduction reaction.
基金supported by the National Natural Science Foundation of China under Grant 61702462the Henan Provincial Science and Technology Research Project under Grants 222102210010 and 222102210064+2 种基金the Research and Practice Project of Higher Education Teaching Reform in Henan Province under Grants 2019SJGLX320 and 2019SJGLX020the Undergraduate Universities Smart Teaching Special Research Project of Henan Province under Grant JiaoGao[2021]No.489-29the Academic Degrees&Graduate Education Reform Project of Henan Province under Grant 2021SJGLX115Y.
文摘Multimodal sentiment analysis aims to understand people’s emotions and opinions from diverse data.Concate-nating or multiplying various modalities is a traditional multi-modal sentiment analysis fusion method.This fusion method does not utilize the correlation information between modalities.To solve this problem,this paper proposes amodel based on amulti-head attention mechanism.First,after preprocessing the original data.Then,the feature representation is converted into a sequence of word vectors and positional encoding is introduced to better understand the semantic and sequential information in the input sequence.Next,the input coding sequence is fed into the transformer model for further processing and learning.At the transformer layer,a cross-modal attention consisting of a pair of multi-head attention modules is employed to reflect the correlation between modalities.Finally,the processed results are input into the feedforward neural network to obtain the emotional output through the classification layer.Through the above processing flow,the model can capture semantic information and contextual relationships and achieve good results in various natural language processing tasks.Our model was tested on the CMU Multimodal Opinion Sentiment and Emotion Intensity(CMU-MOSEI)and Multimodal EmotionLines Dataset(MELD),achieving an accuracy of 82.04% and F1 parameters reached 80.59% on the former dataset.
基金This study was supported by the National Natural Science Foundation of China(Nos.61703058,81873701).
文摘Studies on the integration of cross-modal information with taste perception has been mostly limited to uni-modal level.The cross-modal sensory interaction and the neural network of information processing and its control were not fully explored and the mechanisms remain poorly understood.This mini review investigated the impact of uni-modal and multi-modal information on the taste perception,from the perspective of cognitive status,such as emotion,expectation and attention,and discussed the hypothesis that the cognitive status is the key step for visual sense to exert influence on taste.This work may help researchers better understand the mechanism of cross-modal information processing and further develop neutrally-based artificial intelligent(AI)system.
