检索结果-内蒙古大学图书馆

6th International Conference on intelligent Computing and signal processing (ICSP)

作者： Xin Jiang Peng Zhou Shu Zhao Yufeng Guo School of Computer Science and Technology Anhui University Hefei China Key Laboratory of Intelligent Computing and Signal Processing Anhui University Hefei China

Technological innovation is becoming one of the critical factors in promoting social development all over the world. The vigorous development of patent applications in recent years provides an opportunity to reveal the inherent laws of innovation, but it also puts forward higher requirements for patent mining technology. An essential step in patent text mining is to build a technical portrait for each patent, that is, to identify the technical phrases involved, which can summarize and represent the patent from a technical perspective. Previous technical phrase extraction methods thoroughly used technical phrases' characteristics and the relationship between technical phrases. Regarding our observations, the relationship between patent texts and technical phrases is also essential. Specifically, critical technical phrases are more relevant to the patent text and can be discovered by the attention mechanism. Motivated by this, we propose an unsupervised technical phrase extraction method based on the attention mechanism named UTESC. Self-attention captures the importance of technical phrases in sentences, and cross-attention captures the relevance between technical phrases and patents. Extensive experiments and algorithm comparisons on patent datasets have proven the effectiveness of our algorithm.

关键词：

来源：评论

学校读者我要写书评

暂无评论

A Robust Framework Based on Mutual Information and Contrastive Learning for Multimodal Violence Detection in Videos 23

A Robust Framework Based on Mutual Information and Contrasti...

引用

15th International Conference on Machine Learning and Computing, ICMLC 2023

作者： Xiong, Guoding Zhou, Yongmei Zhou, Dong Lin, Nankai School of Information Science and Technology Guangzhou Key Laboratory of Multilingual Intelligent Processing Guangdong University of Foreign Studies Guangzhou China School of Computer Science and Technology Guangdong University of Technology Guangzhou China

ISBN: (纸本)9781450398411

With the development of modern information, the security of online videos content is receiving more and more *** work often ignores the correlation between different modalities. In order to solve the above problem, the task of violence detection is proposed to identify and locate violent frames in videos. In this paper, we propose a robust framework based on mutual information and contrastive learning for multimodal violence detection in videos. We use mutual information maximization to improve the correlation between audio and visual modalities. Meanwhile, we leverage contrastive learning to control modality fusion to capture modality invariance. We propose a robust framework consisting of three modules: 1) utilize a graph convolutional neural network module to capture global relationships across multiple video frames;2) demonstrate a mutual information module to strengthen the correlation between audio and visual modalities;3) utilize the contrastive learning module to capture the modality invariance to facilitate modality fusion. For comparison with previous studies, we conduct extensive experiments on a public XD-Violence dataset. Experimental results show that our method achieves state-of-the-art performance. © 2023 ACM.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

SCENERY:a lightweight block cipher based on Feistel structure

引用

Frontiers of computer Science 2022年第3期16卷 193-202页

作者： Jingya FENG Lang LI Hunan Provincial Key Laboratory of Intelligent Information Processing and Application Hengyang Normal UniversityHengyang 421002China College of Information Science and Engineering Hunan Normal UniversityChangsha 410081China College of Computer Science and Technology Hengyang Normal UniversityHengyang 421002China

In this paper,we propose a new lightweight block cipher called *** main purpose of SCENERY design applies to hardware and software *** is a 64-bit block cipher supporting 80-bit keys,and its data processing consists of 28 *** round function of SCENERY consists of 84×4 S-boxes in parallel and a 32× 32 binary matrix,and we can implement SCENERY with some basic logic *** hardware implementation of SCENERY only requires 1438 GE based on 0.18 um CMOS technology,and the software implementation of encrypting or decrypting a block takes approximately 1516 clock cycles 0118-bit microcontrollers and 364 clock cycles on 64-bit *** with other encryption algorithms,the performance of SCENERY is well balanced for both hardware and *** the security analyses,SCENERY can achieve enough security margin against known attacks,such as differential cryptanalysis,linear cryptanalysis,impossible differential cryptanalysis and related-key attacks.

关键词： lightweight block cipher feistel structure cryptanalysis internet of things

来源：评论

学校读者我要写书评

暂无评论

PACE: Reputation-Driven Hierarchical Incentive Mechanism for Trusted Healthcare Transactions in Model Marketplace

引用

IEEE Transactions on Consumer Electronics 2024年第1期71卷 1571-1583页

作者： Lu, Jianfeng Dou, Xiaochi Rathore, Bharati Jhaveri, Rutvij H. Seid, Abegaz Mohammed Erbad, Aiman Wuhan University of Science and Technology School of Computer Science and Technology Wuhan430065 China Wuhan University of Science and Technology Hubei Province Key Laboratory of Intelligent Information Processing and Real-time Industrial System China Ministry of Education Key Laboratory of Social Computing and Cognitive Intelligence Dalian University of Technology China United Kingdom Pandit Deendayal Energy University Department of Computer Science and Engineering School of Technology India Hamad Bin Khalifa University College of Science and Engineering Doha Qatar Qatar University College of Engineering Doha Qatar

Federated learning-based model marketplaces have the potential to securely leverage healthcare data for efficient healthcare transactions. However, the willingness to participate in this marketplace is severely hindered by self-interested model sellers, who maximize their personal gains via distorting the quality of their models. To this end, we propose a rePutation-driven hierArchical inCentive mEchanism, called PACE, which combines contract cooperation and game pricing to motivate model sellers to be honest, thereby improving the sustainability of the model marketplace via two-layer cooperation. Specifically, at the training layer, we design a reward payment rule based on multidimensional contracts, while at the trading layer, we use a competitive market pricing game to develop the accuracy pricing rule. In particular, the above rules cleverly combine the fine-grained model sellers' reputations with the willingness of healthcare devices to indirectly affect the model sellers' utilities. Theoretically, we prove that PACE can effectively incentivize model sellers to truthfully report model quality while satisfying both individual rationality and incentive compatibility with a fairness guarantee. Experimental results conducted on four real-world datasets show that our PACE improves the global performance by up to 12.85% and reduces the global loss by at least 19.23% compared with the state-of-the-art baselines. © 1975-2011 IEEE.

关键词： Marketplaces

来源：评论

学校读者我要写书评

暂无评论

Auto-Enhanced Population Diversity with Two Options 1

引用

14th International Symposium on Intelligence Computation and Applications, ISICA 2023

作者： Ou, Yangcong Yang, Ming Guan, Jing School of Computer Science China University of Geosciences Wuhan430078 China Hubei Key Laboratory of Intelligent Geo-Information Processing China University of Geosciences Wuhan430078 China China Ship Development and Design Center Wuhan430064 China

ISBN: (数字)9789819743933

ISBN: (纸本)9789819743926

This paper proposes an auto-enhanced population diversity with two options (TO-AEPD) for addressing the premature convergence of CMA-ES. TO-AEPD is founded on a modification of AEPD. By quantifying the population distribution in each dimension, TO-AEPD allows the identification of population convergence or stagnation instances. When convergence or stagnation is identified, the population must be diversified to acceptable levels. Two population diversification enhancement methods are used to balance exploration and exploitation. TO-AEPD selects the population diversification strategy with exploration when the current round of optimization does not produce a new optimal solution or when convergence accuracy is achieved. Otherwise, population diversification with an exploitation strategy is implemented. TO-AEPD-CMAES describes the CMAES that incorporates TO-AEPD. Experimentation demonstrated that TO-AEPD-CMAES outperforms several competing algorithms on benchmarks, significantly enhancing the algorithm’s performance and validating the model’s viability and efficacy. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024.

关键词： Optimization

来源：评论

学校读者我要写书评

暂无评论

An Optimization Method Based on Deep Learning for Electromagnetic Inverse Scattering Problems

An Optimization Method Based on Deep Learning for Electromag...

引用

International Applied Computational Electromagnetics Society Symposium (ACES)

作者： Wangchao Hu Guoda Xie Jinpeng Dai Guilin Hou Xiang Li Zhixiang Huang The Key Laboratory of Intelligent Computing and Signal Processing Anhui University Hefei China Education of Anhui Province Anhui Province Key Laboratory of Target Recognition and Feature Extraction Luan China The Key Laboratory of Intelligent Computing and Signal Processing Ministry of Education Hefei China Key Laboratory of Electromagnetic Environmental Sensing Anhui Province Anhui University Hefei China Anhui University Hefei China

ISBN: (数字)9798350355581

ISBN: (纸本)9798350355598

In this paper, a deep learning optimization method combining U-net model and cycle generative adversarial network (Cycle-GAN) is proposed to efficiently solve the electromagnetic inverse scattering (EMIS) problems. Firstly, a U-net model is used to obtain a rough preliminary reconstruction. Through down-sampling and up-sampling processes, the basic features of the object are extracted and reconstructed from the measured scattered field, providing a coarse dielectric parameter distribution. Then, a Cycle-GAN model further refines the output of the U-net. The Cycle-GAN model, through its adversarial training mechanism involving a generator and a discriminator, enhances the details of the preliminary reconstruction, resulting in a finely detailed dielectric parameter distribution his method improves the problem of rough results obtained by using U-net method alone, and can effectively solve and optimize EMIS problems.

关键词： Deep learning Inverse problems Computational modeling Electromagnetic interference Electromagnetic scattering Optimization methods Feature extraction Dielectric measurement Electromagnetics Image reconstruction

来源：评论

学校读者我要写书评

暂无评论

Extending generalized unsupervised manifold alignment

引用

Science China(Information Sciences) 2022年第7期65卷 139-156页

作者： Xiaoyi YIN Zhen CUI Hong CHANG Bingpeng MA Shiguang SHAN Key Laboratory of Intelligent Information Processing of Chinese Academy of Sciences (CAS) Institute of Computing Technology Chinese Academy of Sciences University of Chinese Academy of Sciences School of Computer Science and Engineering Nanjing University of Science and Technology CAS Center for Excellence in Brain Science and Intelligence Technology

Building connections between different data sets is a fundamental task in machine learning and related application community. With proper manifold alignment, the correspondences between data sets will assist us with comprehensive study of data processes and analyses. Despite the several progresses in semi-supervised and unsupervised scenarios, potent manifold alignment methods in generalized and realistic circumstances remain in absence. Besides, theretofore unsupervised algorithms seldom prove themselves mathematically. In this paper, we devise an efficient method to properly solve the unsupervised manifold alignment problem and denominate it as extending generalized unsupervised manifold alignment(EGUMA)method. More specifically, an explicit relaxed integer programming method is adopted to solve the unsupervised manifold alignment problem, which reconciles three factors covering the updated local structure matching, the the feature comparability and geometric preservation. An additional effort is retained on extending the Frank Wolfe algorithm to tacking our optimization problem. Besides our previous endeavors we adopt a new strategy for neighborhood discovery in the manifolds. The main advantages over previous methods accommodate(1) simultaneous alignment and discovery of manifolds;(2) complete unsupervised learning structure without any prerequisite correspondence;(3) more concise local geometry for the embedding space;(4) efficient alternative optimization;(5) strict mathematical analysis on the convergence and efficiency issues. Experiments on real-world applications verify the high accuracy and efficiency of our proposed method.

关键词： unsupervised manifold alignment Unsupervised Manifold Alignment relaxed integer programming Relaxed Integer Programming Frank Wolfe algorithm Frank Wolfe Algorithm

来源：评论

学校读者我要写书评

暂无评论

Side-channel analysis attacks based on deep learning network

引用

Frontiers of computer Science 2022年第2期16卷 35-45页

作者： Yu OU Lang LI Hunan Provincial Key Laboratory of Intelligent Information Processing and Application Hengyang Normal UniversityHengyang 421002China College of Information Science and Engineering Hunan Normal UniversityChangsha 410081China College of Computer Science and Technology Hengyang Normal UniversityHengyang 421002China

There has been a growing interest in the sidechannel analysis(SCA)field based on deep learning(DL)*** DL network or model has been developed to improve the efficiency of ***,few studies have investigated the impact of the different models on attack results and the exact relationship between power consumption traces and intermediate *** on the convolutional neural network and the autoencoder,this paper proposes a Template Analysis Pre-trained DL Classification model named TAPDC which contains three *** TAPDC model detects the periodicity of power trace,relating power to the intermediate values and mining the deeper features by the multi-layer convolutional *** implement the TAPDC model and compare it with two classical models in a fair *** evaluative results show that the TAPDC model with autoencoder and deep convolution feature extraction structure in SCA can more effectively extract information from power consumption ***,Using the classifier layer,this model links power information to the probability of intermediate *** completes the conversion from power trace to intermediate values and greatly improves the efficiency of the power attack.

关键词： side-channel analysis template attack machine learning deep learning

来源：评论

学校读者我要写书评

暂无评论

Research on Tibetan-Chinese Neural Machine Translation Integrating Statistical Method 6

Research on Tibetan-Chinese Neural Machine Translation Integ...

引用

6th International Conference on Machine Learning and Natural Language processing, MLNLP 2023

作者： Zhou, Maoxian The Computer College of Qinghai Normal University Qinghai Xining810016 China Tibetan Information Processing and Machine Translation Key Laboratory of Qinghai Province Qinghai Xining810008 China Qinghai Engineering and Technology Research Center for Tibetan Information Processing Qinghai Xining810008 China The State Key Laboratory of Tibetan Intelligent Information Processing and Application Qinghai Xining810008 China

ISBN: (纸本)9798400709241

In recent years, with the emergence of deep learning methods, Neural Machine Translation has become a new research direction of machine translation. Due to the scarcity of digital resources in Tibetan, there is only a small-scale Tibetan-Chinese bilingual parallel corpus to train translation models. In the case of insufficient resources, the performance of Neural Machine Translation decreases significantly. In this regard, considering that the word alignment results in Tibetan-Chinese Statistical Machine Translation are better, and the alignment information in Tibetan-Chinese Neural Machine Translation model is significantly different from that in Tibetan-Chinese Statistical Machine Translation, this paper proposes a Tibetan-Chinese Neural Machine Translation method that integrates statistical methods. We using Statistical Machine Translation method to generate word alignment information of Tibetan-Chinese parallel corpus to supervise the training process of Tibetan-Chinese Neural Machine Translation model to make the model achieve more accurate translation and alignment effect. Experimental results show that this method can effectively improve translation quality in low resource environment. © 2023 ACM.

关键词： Neural machine translation

来源：评论

学校读者我要写书评

暂无评论

Interpretable multi-view attention network for drug-drug interaction prediction

Interpretable multi-view attention network for drug-drug int...

引用

2023 IEEE International Conference on Bioinformatics and Biomedicine, BIBM 2023

作者： Lin, Xuan Wen, Qi Yang, Sijie Yu, Zu-Guo Long, Yahui Zeng, Xiangxiang Xiangtan University College of Computer Science China Xiangtan University Key Laboratory of Intelligent Computing and Information Processing of Ministry of Education China Singapore Hunan University College of Information Science and Engineering China

ISBN: (纸本)9798350337488

Drug-drug interaction (DDI) plays an increasingly crucial role in drug discovery. Predicting potential DDI is also essential for clinical research. Given the high cost and risk of wet-lab experiments, in-silico DDI prediction is an alternative choice. Recently, deep learning methods have been developed for DDI prediction. However, most of existing methods focus on feature extraction from either molecular SMILES sequences or drug interactive networks, ignoring the valuable complementary information that can be derived from these two views. In this paper, we propose a novel interpretable Multi-View Attention network (MVA-DDI) for DDI prediction. MVA-DDI can effectively extracts drug representations from different perspectives to improve DDI prediction. Specifically, for a given drug, we design a transformer-based encoder and a graph convolutional networkbased encoder to learn sequence and graph representations from SMILES sequence and molecular graph, respectively. To fully exploit the complementary information between the sequence and molecular views, an attention mechanism is further adopted to adaptively aggregate the sequence and graph representations by taking the importance of different views into accounts, generating the final drug representations. Comparison experiments demonstrated that our MVA-DDI1 model achieved superior performance to state-of-the-art models on DDI prediction. © 2023 IEEE.

关键词： Contrastive learning Drug-drug interaction prediction Interpretable attention network Multi-view learning

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：