Constructing an effective common latent embedding by aligning the latent spaces of cross-modal variational autoencoders（VAEs） is a popular strategy for generalized zero-shot learning（GZSL）. However, due to the lac...
详细信息
Constructing an effective common latent embedding by aligning the latent spaces of cross-modal variational autoencoders（VAEs） is a popular strategy for generalized zero-shot learning（GZSL）. However, due to the lack of fine-grained instance-wise annotations, existing VAE methods can easily suffer from the posterior collapse problem. In this paper, we propose an innovative asymmetric VAE network by aligning enhanced feature representation（AEFR） for GZSL. Distinguished from general VAE structures, we designed two asymmetric encoders for visual and semantic observations and one decoder for visual reconstruction. Specifically, we propose a simple yet effective gated attention mechanism（GAM） in the visual encoder for enhancing the information interaction between observations and latent variables, alleviating the possible posterior collapse problem effectively. In addition, we propose a novel distributional decoupling-based contrastive learning（D2-CL） to guide learning classification-relevant information while aligning the representations at the taxonomy level in the latent representation space. Extensive experiments on publicly available datasets demonstrate the state-of-the-art performance of our method. The source code is available at https://***/seeyourmind/AEFR.
A multi-strategy hybrid whale optimization algorithm(MSHWOA)for complex constrained optimization problems is proposed to overcome the drawbacks of easily trapping into local optimum,slow convergence speed and low opti...
详细信息
A multi-strategy hybrid whale optimization algorithm(MSHWOA)for complex constrained optimization problems is proposed to overcome the drawbacks of easily trapping into local optimum,slow convergence speed and low optimization ***,the population is initialized by introducing the theory of good point set,which increases the randomness and diversity of the population and lays the foundation for the global optimization of the ***,a novel linearly update equation of convergence factor is designed to coordinate the abilities of exploration and *** the same time,the global exploration and local exploitation capabilities are improved through the siege mechanism of Harris Hawks optimization ***,the simulation experiments are conducted on the 6 benchmark functions and Wilcoxon rank sum test to evaluate the optimization performance of the improved *** experimental results show that the proposed algorithm has more significant improvement in optimization accuracy,convergence speed and robustness than the comparison algorithm.
The coalescence and missed detection are two key challenges in Multi-Target Tracking(MTT).To balance the tracking accuracy and real-time performance,the existing Random Finite Set(RFS)based filters are generally diffi...
详细信息
The coalescence and missed detection are two key challenges in Multi-Target Tracking(MTT).To balance the tracking accuracy and real-time performance,the existing Random Finite Set(RFS)based filters are generally difficult to handle the above problems simultaneously,such as the Track-Oriented marginal Multi-Bernoulli/Poisson(TOMB/P)and Measurement-Oriented marginal Multi-Bernoulli/Poisson(MOMB/P)*** on the Arithmetic Average(AA)fusion rule,this paper proposes a novel fusion framework for the Poisson Multi-Bernoulli(PMB)filter,which integrates both the advantages of the TOMB/P filter in dealing with missed detection and the advantages of the MOMB/P filter in dealing with *** order to fuse the different PMB distributions,the Bernoulli components in different Multi-Bernoulli(MB)distributions are associated with each other by Kullback-Leibler Divergence(KLD)***,an adaptive AA fusion rule is designed on the basis of the exponential fusion weights,which utilizes the TOMB/P and MOMB/P updates to solve these difficulties in ***,by comparing with the TOMB/P and MOMB/P filters,the performance of the proposed filter in terms of accuracy and efficiency is demonstrated in three challenging scenarios.
Multi-target tracking is facing the difficulties of modeling uncertain motion and observation *** tracking algorithms are limited by specific models and priors that may mismatch a real-world *** this paper,considering...
详细信息
Multi-target tracking is facing the difficulties of modeling uncertain motion and observation *** tracking algorithms are limited by specific models and priors that may mismatch a real-world *** this paper,considering the model-free purpose,we present an online Multi-Target Intelligent Tracking(MTIT)algorithm based on a Deep Long-Short Term Memory(DLSTM)network for complex tracking requirements,named the MTIT-DLSTM ***,to distinguish trajectories and concatenate the tracking task in a time sequence,we define a target tuple set that is the labeled Random Finite Set(RFS).Then,prediction and update blocks based on the DLSTM network are constructed to predict and estimate the state of targets,***,the prediction block can learn the movement trend from the historical state sequence,while the update block can capture the noise characteristic from the historical measurement ***,a data association scheme based on Hungarian algorithm and the heuristic track management strategy are employed to assign measurements to targets and adapt births and *** results manifest that,compared with the existing tracking algorithms,our proposed MTIT-DLSTM algorithm can improve effectively the accuracy and robustness in estimating the state of targets appearing at random positions,and be applied to linear and nonlinear multi-target tracking scenarios.
Rechargeable Mg-ion batteries(MIBs)have attracted much more attentions by virtue of the high capacity from the two electrons ***,the reversible Mg^(2+)diffusion in cathode materials is restricted by the strong interac...
详细信息
Rechargeable Mg-ion batteries(MIBs)have attracted much more attentions by virtue of the high capacity from the two electrons ***,the reversible Mg^(2+)diffusion in cathode materials is restricted by the strong interactions between the high-polarized bivalent Mg^(2+)ions and anionic ***,we design and propose a hetero-structural VO_(2)(R)-VS_(4)cathode,in which the re-delocalized d-electrons can effectively shield the polarity of Mg^(2+)***,the electrons should spontaneously transfer from VS_(4)to VO_(2)(R)through the interfaces of hetero-structure due to the lower work function value of VS_(4).Furthermore,the internal electrons transfer lead to the electronic injection into VO_(2)(R)from VS_(4)and the partially broken V-V dimers,indicating the presence of lone pair electrons and charge *** from the shield effect of re-delocalized electrons,and the weakened attraction between cations and O/S anions enables more S^(2-)-S_(2)^(2-)redox groups to participate the electrochemical reactions and compensate the double charge of Mg^(2+)***,VO_(2)(R)-VS_(4)hetero-structure exhibits a high specific capacity of 554 mA h g^(-1)at 50 mA g^(-1).It is believed that the charge re-delocalization of cathode extremely boost the Mg^(2+)ions migration for the high-capacity of MIBs.
The energy consumption of heating,ventilation,and air conditioning(HVAC)systems holds a significant position in building energy usage,accounting for about 65%of the total energy ***,with the advancement of building au...
详细信息
The energy consumption of heating,ventilation,and air conditioning(HVAC)systems holds a significant position in building energy usage,accounting for about 65%of the total energy ***,with the advancement of building automation,the energy consumption of ventilation systems continues to *** study focuses on improving the performance of spherical tuyeres in HVAC *** primarily utilizes neural networks and multi-island genetic algorithms(MIGA)for multi-parameter *** employing methods such as structural parameterization,accurate and fast computational fluid dynamics(CFD)simulations,a minimized sample space,and a rational optimization strategy,the time cycle of the optimization process is ***,a new comprehensive evaluation index is proposed in this research to describe the performance of spherical tuyeres,which can be used to more accurately assess spherical tuyeres with different *** results show that by establishing a neural network prediction model and combining it with the multi-island genetic algorithm,a novel spherical tuyere design was successfully *** optimized novel spherical tuyeres achieved a 27.05%reduction in the spherical tuyeres effective index(STEI)compared to the traditional spherical ***,the resistance decreased by 15.68%,and the jet length increased by 7.57%.The experimental results demonstrate that our proposed optimization method exhibits high accuracy,good generalization capability,and excellent agreement at different Reynolds numbers.
In recent years,the demand for real-time data processing has been increasing,and various stream processing systems have *** the amount of data input to the stream processing system fluctuates,the computing resources r...
详细信息
In recent years,the demand for real-time data processing has been increasing,and various stream processing systems have *** the amount of data input to the stream processing system fluctuates,the computing resources required by the stream processing job will also *** resources used by stream processing jobs need to be adjusted according to load changes,avoiding the waste of computing *** present,existing works adjust stream processing jobs based on the assumption that there is a linear relationship between the operator parallelism and operator resource consumption(e.g.,throughput),which makes a significant deviation when the operator parallelism *** paper proposes a nonlinear model to represent operator *** divide the operator performance into three stages,the Non-competition stage,the Non-full competition stage,and the Full competition *** our proposed performance model,given the parallelism of the operator,we can accurately predict the CPU utilization and operator *** with actual experiments,the prediction error of our model is below 5%.We also propose a quick accurate auto-scaling(QAAS)method that uses the operator performance model to implement the auto-scaling of the operator parallelism of the Flink *** to previous work,QAAS is able to maintain stable job performance under load changes,minimizing the number of job adjustments and reducing data backlogs by 50%.
Heterogeneous photocatalysts exhibit high catalytic efficiency in the degradation of pollutants,but their stability and repeatability is not very good and requires high structural *** by nanosizing the pure Bi_(2)WO_(...
详细信息
Heterogeneous photocatalysts exhibit high catalytic efficiency in the degradation of pollutants,but their stability and repeatability is not very good and requires high structural *** by nanosizing the pure Bi_(2)WO_(6)(BWO)photocatalyst without constructing a heterojunction,there is a significant improvement in its performance,with an enhancement effect of about 2.3 times(99.43%).The high photocatalytic degradation efficiency of the material can be attributed to the enhanced light absorption effect brought by the three-dimensional inverse-opal structure SiO_(2)(IS)and the abundant surface adsorbed oxygen generated after the formation of Si–O–W *** addition,the introduction of IS greatly increases the surface area of nanostructured BWO,which accelerates the charge transfer process,while the adsorbed oxygen promotes the participation of·O^(2−) in the photocatalytic reaction,thereby accelerating the consumption of photo-generated electrons and ultimately improving the separation of charge ***,the matched photonic bandgap further improves the absorption and utilization of light of the *** this work,we constructs Si–O–W bonds to obtain inverse-opal SiO_(2)/Bi_(2)WO_(6) with uniformly growth of pure phase nano BWO,which provides a feasible strategy for the preparation of high-performance pure-phase photocatalysts.
Sharding technology achieves parallel processing of transactions by dividing the network into multiple independent parts, namely shards, significantly increasing the throughput of the blockchain system and reducing tr...
详细信息
Naturally optimized successful synovial joints with lightweight,high load-carrying,ultra-low friction and wear have attracted tribological communities to constantly imitate and *** impressive advances in cartilage lub...
详细信息
Naturally optimized successful synovial joints with lightweight,high load-carrying,ultra-low friction and wear have attracted tribological communities to constantly imitate and *** impressive advances in cartilage lubrication,extending such extraordinary performance advantages to macroscale solid lubrication remains a ***,inspired by the fascinating interplay of synovial joints,a novel kind of trans-scale hierarchical structured ceramic-based composite was ***-ducing microscale Ag microspheres(AgMs)“cartilage”layer and nanoscale Ag quantum dots/MXene(AgQDs/MXene)“synovial fluid”into the interior and exterior of printed macroscale SiOC“hard bone”realistically restores the gradient structure of synovial joint *** resulted composite with ideal compressive strength(70.44 MPa)can achieve a 60.53%friction reduction and a low wear rate(2.05×10^(−6)mm 3 N^(−1)m^(−1))in dry tribo-contact for 3600 sliding cycles,while also maintaining considerable low friction(∼0.11)over 10,000 sliding cycles and long-term stable lubrication(∼0.13)for up to 50,000 re-ciprocating *** extraordinary performance can be explained by the division of macro contacts,full loading of AgMs and AgQDs/MXene,abrasive debris capture and removal,as well as the shear rolling effect induced by friction *** work opens a new avenue to develop structural lubricating mate-rials for complex engineering applications.
暂无评论