检索结果-南通市图书馆

Conditional selection with CNN augmented transformer for multimodal affective analysis

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

CAAI Transactions on Intelligence technology 2024年第4期9卷 917-931页

作者： Jianwen Wang Shiping Wang Shunxin Xiao Renjie Lin Mianxiong Dong Wenzhong Guo College of Computer and Data Science Fuzhou UniversityFuzhouChina College of Computer and Cyber Security Fujian Normal UniversityFuzhouChina Key Laboratory of Network Computing and Intelligent Information Processing Fuzhou UniversityFuzhouChina Digital Fujian institute of Big Data Security technology Fujian Normal UniversityFuzhouChina Department of Sciences and Informatics Muroran Institute of TechnologyMuroranJapan

Attention mechanism has been a successful method for multimodal affective analysis in recent years. Despite the advances, several significant challenges remain in fusing language and its nonverbal context information. One is to generate sparse attention coefficients associated with acoustic and visual modalities, which helps locate critical emotional se-mantics. The other is fusing complementary cross‐modal representation to construct optimal salient feature combinations of multiple modalities. A Conditional Transformer Fusion Network is proposed to handle these problems. Firstly, the authors equip the transformer module with CNN layers to enhance the detection of subtle signal patterns in nonverbal sequences. Secondly, sentiment words are utilised as context conditions to guide the computation of cross‐modal attention. As a result, the located nonverbal fea-tures are not only salient but also complementary to sentiment words directly. Experi-mental results show that the authors’ method achieves state‐of‐the‐art performance on several multimodal affective analysis datasets.

关键词： affective computing data fusion information fusion multimodal approaches

An intelligent active probing and trace-back scheme for IoT anomaly detection

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Digital Communications and Networks 2024年第1期10卷 168-181页

作者： Luying Wang Lingyi Chen Neal N.Xiong Anfeng Liu Tian Wang Mianxiong Dong School of Computer Science and Engineering Central South UniversityChangsha 410083China Department of Computer Science and Mathematics Sul Ross State UniversityAlpineTX 79830USA Artificial Intelligence and Future Networks Beijing Normal University&UICZhuhai 519087China Department of Information and Electronic Engineering Muroran Institute of TechnologyMuroran 050-8585Japan

Due to their simple hardware,sensor nodes in IoT are vulnerable to attack,leading to data routing blockages or malicious tampering,which significantly disrupts secure data *** Intelligent Active Probing and Trace-back Scheme for IoT Anomaly Detection(APTAD)is proposed to collect integrated IoT data by recruiting Mobile Edge Users(MEUs).(a)An intelligent unsupervised learning approach is used to identify anomalous data from the collected data by MEUs and help to identify anomalous nodes.(b)Recruit MEUs to trace back and propose a series of trust calculation methods to determine the trust of nodes.(c)The last,the number of active detection packets and detection paths are designed,so as to accurately identify the trust of nodes in IoT at the minimum cost of the network.A large number of experimental results show that the recruiting cost and average anomaly detection time are reduced by 6.5 times and 34.33%respectively,while the accuracy of trust identification is improved by 20%.

关键词： Anomaly detection Internet of things Integrating data collection Mobile edge users Intelligent

Deep Guided Attention Network for Joint Denoising and Demosaicing in Real Image

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Chinese Journal of Electronics 2024年第1期33卷 303-312页

作者： Tao ZHANG Ying FU Jun ZHANG School of Computer Science and technology Beijing Institute of Technology Advanced Reasearch institute of Multidisciplinary Science Beijing Institute of Technology

Denoising(DN) and demosaicing(DM) are the first crucial stages in the image signal processing pipeline. Recently, researches pay more attention to solve DN and DM in a joint manner, which is an extremely undetermined inverse problem. Existing deep learning methods learn the desired prior on synthetic dataset, which limits the generalization of learned network to the real world data. Moreover, existing methods mainly focus on the raw data property of high green information sampling rate for DM, but occasionally exploit the high intensity and signalto-noise(SNR) of green channel. In this work, a deep guided attention network(DGAN) is presented for real image joint DN and DM(JDD), which considers both high SNR and high sampling rate of green information for DN and DM, respectively. To ease the training and fully exploit the data property of green channel, we first train DN and DM sub-networks sequentially and then learn them jointly, which can alleviate the error accumulation. Besides, in order to support the real image JDD, we collect paired raw clean RGB and noisy mosaic images to conduct a realistic dataset. The experimental results on real JDD dataset show the presented approach performs better than the state-of-the-art methods, in terms of both quantitative metrics and qualitative visualization.

关键词： Training Visualization Inverse problems Green products Noise reduction Pipelines Signal processing

Fast Cross-Platform Binary Code Similarity Detection Framework Based on CFGs Taking Advantage of NLP and Inductive GNN

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Chinese Journal of Electronics 2024年第1期33卷 128-138页

作者： Jinxue PENG Yong WANG Jingfeng XUE Zhenyan LIU School of Computer Science and technology Beijing Institute of Technology

Cross-platform binary code similarity detection aims at detecting whether two or more pieces of binary code are similar or not. Existing approaches that combine control flow graphs(CFGs)-based function representation and graph convolutional network(GCN)-based similarity analysis are the best-performing ones. Due to a large amount of convolutional computation and the loss of structural information, the use of convolution networks will inevitably bring problems such as high overhead and sometimes inaccuracy. To address these issues, we propose a fast cross-platform binary code similarity detection framework that takes advantage of natural language processing(NLP)and inductive graph neural network(GNN) for basic blocks embedding and function representation respectively by simulating extracting structural features and temporal features. GNN’s node-centric and small batch is a suitable training way for large CFGs, it can greatly reduce computational overhead. Various NLP basic block embedding models and GNNs are evaluated. Experimental results show that the scheme with long short term memory(LSTM)for basic blocks embedding and inductive learning-based Graph SAGE(GAE) for function representation outperforms the state-of-the-art works. In our framework, we can take only 45% overhead. Improve efficiency significantly with a small performance trade-off.

关键词： Training Semantics Process control Binary codes Feature extraction Natural language processing Graph neural networks

同方期刊数据库

Prescribed-time leader-following consensus of linear multi-agent systems by bounded linear time-varying protocols

在线全文

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2024年第1期67卷 140-153页

作者： Kai ZHANG Bin ZHOU Xuefei YANG Guangren DUAN Center for Control Theory and Guidance technology Harbin Institute of Technology

This paper considers the prescribed-time leader-following consensus problem of input-constrained linear multi-agent systems under generally directed communication topology in two cases:the Laplacian matrix related to the entire communication topology between agents is either known or *** particular,the consensus problem for the former case is solved by a novel bounded linear time-varying(LTV) protocol,where the feedback gain is formulated by the parametric Lyapunov equation and the knowledge of the Laplacian ***,by utilizing a distributed observer,a fully bounded LTV protocol is proposed for the latter *** should be noted that,compared with the existing results,the system under consideration is more general,the designed protocols are linear,and the consensus problem is accomplished even in a fully distributed ***,the effectiveness of the proposed approach is verified by a numerical example.

关键词： leader-following consensus time-vary feedback prescribed-time control fully distributed consensus

Prescribed time control based on the periodic delayed sliding mode surface without singularities

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2024年第7期67卷 256-270页

作者： Bin ZHOU Yi DING Kang-Kang ZHANG Guang-Ren DUAN Center for Control Theory and Guidance technology Harbin Institute of Technology

In this paper, nonsingular prescribed-time control is studied based on periodic delayed sliding mode surfaces. Different from the existing sliding mode control, where either singularity problems may appear or the convergence time depends on the initial state, the proposed sliding mode control approaches can achieve prescribed-time convergence without singularity. The proposed nonsingular sliding mode control approaches can be applied to both second-order and high-order nonlinear systems with prescribed-time convergence. As applications of the proposed sliding mode control approaches, the control of hypersonic vehicle systems is revisited. Numerical simulations on the nonlinear model of the hypersonic vehicle system show the effectiveness of the proposed methods.

关键词： prescribed-time control smooth periodic delayed feedback sliding mode control high-order nonlinear systems hypersonic vehicle systems

同方期刊数据库

A model reduction approach for discrete-time linear time-variant systems with delayed inputs

在线全文

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2024年第4期67卷 197-217页

作者： Ai-Guo WU Guang-Ren DUAN Yu WANG Jie ZHANG School of Mechanical Engineering and Automation Harbin Institute of Technology(Shenzhen) HIT Wuhu Robot technology Research institute Center for Control Science and technology Southern University of Science and Technology

A model reduction approach is presented for discrete-time linear time-variant input-delayed systems. According to this proposed approach, a dynamical variable is constructed by taking advantage of the current state and historical information of input. It is revealed that the behavior of this dynamical variable is governed by a discrete-time linear delay-free system. It is worth noting that the presented variable transformation does not require the system matrix to be invertible. Based on the reduced delay-free models,stabilizing control laws can be easily obtained for the original delayed system. For the case with a single input delay, the constructed variable is an exact prediction for the future state, and thus the stabilizing control law could be designed by replacing the future state with its prediction. Finally, three discrete-time periodic systems with delayed input are employed to illustrate how to utilize the presented model reduction approaches.

关键词： reduction approach predictor-based feedback time-varying systems delayed input periodic system

同方期刊数据库

Ru-doped functional porous materials for electrocatalytic water splitting

在线全文

学校读者我要写书评

暂无评论

Nano Research 2024年第3期17卷 982-1002页

作者： Chongao Tian Rui Liu Yu Zhang Wenxiu Yang Bo Wang Beijing Key Laboratory of Photoelectronic/Electrophotonic Conversion Materials Advanced Technology Research Institute(Jinan)Advanced Research Institute of Multidisciplinary ScienceSchool of Chemistry and Chemical EngineeringBeijing Institute of TechnologyBeijing 100081China Beijing institute of technology Library Beijing Institute of TechnologyBeijing 100081China

Electrolytic water splitting(EWS)is an attractive and promising technique for the production of hydrogen ***,the sluggish kinetic rate of hydrogen/oxygen evolution reactions leads to a high overpotential and low energy *** to date,Pt/Ir-based nanocatalysts have become the state-of-the-art EWS catalysts,but disadvantages such as high cost and low earth abundance greatly limit their applications in EWS *** an attractive candidate for the Pt/Ir catalysts,series of Ru-based nanomaterials have aroused much attention for their low price,Pt-like hydrogen bond strength,and high EWS *** particular,Ru-doped functional porous materials have been becoming one of the most representative EWS catalysts,which can not only achieve the dispersion and adjustment for active Ru sites,but also simultaneously solve the problems of mass transfer and catalytic conversion in *** this review,the design and preparation strategies of Ru-doped functional porous materials toward EWS in recent years are summarized,including Ru-doped metal organic frameworks(MOFs),Ru-doped porous organic polymers(POPs),and their ***,detailed structure–activity relationships induced by the tuned geometric/electronic structures of Ru-doped functional porous materials are further depicted in this *** but not least,the challenges and perspectives of Ru-doped functional porous materials catalysts are reasonably proposed to provide fresh ideas for the design of Ru-based EWS catalysts.

关键词： Ru-based catalyst functional porous materials electrolytic water splitting oxygen evolution reaction hydrogen evolution reaction

An ultra-deep TSV technique enabled by the dual catalysis-based electroless plating of combined barrier and seed layers

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Microsystems & Nanoengineering 2024年第3期10卷 409-419页

作者： Yuwen Su Yingtao Ding Lei Xiao Ziyue Zhang Yangyang Yan Zhifang Liu Zhiming Chen Huikai Xie School of Integrated Circuits and Electronics Beijing Institute of TechnologyBeijingChina Chongqing institute of Microelectronics and Microsystems Beijing Institute of TechnologyChongqingChina

Silicon interposers embedded with ultra-deep through-silicon vias(TSVs)are in great demand for the heterogeneous integration and packaging of opto-electronic chiplets and microelectromechanical systems(MEMS)*** the cost-effective and reliable manufacturing of ultra-deep TSVs,the formation of continuous barrier and seed layers remains a crucial challenge to ***,we present a novel dual catalysis-based electroless plating(ELP)technique by tailoring polyimide(PI)liner surfaces to fabricate dense combined Ni barrier/seed layers in ultradeep *** additional to the conventional acid catalysis procedure,a prior catalytic step in an alkaline environment is proposed to hydrolyze the PI surface into a polyamide acid(PAA)interfacial layer,resulting in additional catalysts and the formation of a dense Ni layer that can function as both a barrier layer and a seed layer,particularly at the bottom of the deep *** with depths larger than 500μm and no voids are successfully fabricated in this *** fabrication process involves low costs and *** a fabricated 530-μm-deep TSV with a diameter of 70μm,the measured depletion capacitance and leakage current are approximately 1.3 pF and 1.7 pA at 20 V,respectively,indicating good electrical *** proposed fabrication strategy can provide a cost-effective and feasible solution to the challenge of manufacturing ultra-deep TSVs for modern 3D heterogeneous integration and packaging applications.

关键词： properties. catalysis alkaline