检索结果-内蒙古大学图书馆

Paradigm Shift in Natural Language processing

Machine Intelligence Research 2022年第3期19卷 169-183页

作者： Tian-Xiang Sun Xiang-Yang Liu Xi-Peng Qiu Xuan-Jing Huang School of Computer Science Fudan UniversityShanghai 200438China Shanghai Key Laboratory of Intelligent Information Processing Fudan UniversityShanghai 200438China

In the era of deep learning, modeling for most natural language processing (NLP) tasks has converged into several mainstream paradigms. For example, we usually adopt the sequence labeling paradigm to solve a bundle of tasks such as POS-tagging, named entity recognition (NER), and chunking, and adopt the classification paradigm to solve tasks like sentiment analysis. With the rapid progress of pre-trained language models, recent years have witnessed a rising trend of paradigm shift, which is solving one NLP task in a new paradigm by reformulating the task. The paradigm shift has achieved great success on many tasks and is becoming a promising way to improve model performance. Moreover, some of these paradigms have shown great potential to unify a large number of NLP tasks, making it possible to build a single model to handle diverse tasks. In this paper, we review such phenomenon of paradigm shifts in recent years, highlighting several paradigms that have the potential to solve different NLP tasks.

关键词： Natural language processing pre-trained language models deep learning sequence-to-sequence paradigm shift

来源：评论

学校读者我要写书评

暂无评论

Towards Disturbance Rejection in Feature Pyramid Network

IEEE Transactions on Artificial Intelligence

引用

IEEE Transactions on Artificial Intelligence 2023年第4期4卷 946-958页

作者： Hu, Xiaobo Xu, Wei Gan, Yi Su, Jianbo Zhang, Jun Shanghai Jiao Tong University Joint Institute of UM-SJTU Shanghai200240 China Shanghai Jiao Tong University Research Center of Intelligent Robotics Key Laboratory of System Control and Information Processing Ministry of Education Department of Automation Shanghai200240 China

The multiscale problem is one of the challenging topics in the object detection field. Among the current studies toward this problem, feature pyramid network (FPN) has been shown a superior performance. The core principle of FPN is a mapping for the detectable objects into hierarchy features. These features are designed with different size of receptive fields, which brings good scale invariance in the classification and localization subtasks. However, a disturbance of information dilution is generated by the assignment for the detectable objects into different features. This dilution disturbance is associated with the convolution or pooling operation during the hierarchy feature extraction process. Both of the imprecise localization information in the top features and the fuzzy classification information in the bottom ones limit the performance of FPN. Aiming at the disturbance rejection problem of information dilution, this article presents a feature compensation method. A cross compensation between the semantic information and the localization information in hierarchy features is used to enhance the vanilla FPN. Then, the evaluation metrics in an object detection task are examined. Not only the overall performance, but also the respective metrics for multiscale objects validate a considerable improvement of the FPN with a disturbance compensation mechanism. The experimental results demonstrate an effectiveness of the proposed method in this article. © 2020 IEEE.

关键词： Object detection

来源：评论

学校读者我要写书评

暂无评论

MAGPNet: Multi-Domain Attention-Guided Pyramid Network for Infrared Small Object Detection

引用

Journal of Shanghai Jiaotong University (Science) 2024年 1-18页

作者： Ding, Leqi Wang, Biyun Yao, Lixiu Cai, Yunze Department of Automation Shanghai Jiao Tong UniversityKey Laboratory of System Control and Information Processing of Ministry of Education Shanghai200240 China National Key Laboratory of Air-based Information Perception and Fusion Henan Luoyang471009 China Shanghai Institute of Aerospace Control Technology Infrared Detection Technology R&D Center of China Aerospace Science and Technology Corporation Shanghai201109 China

To overcome the obstacles of poor feature extraction and little prior information on the appearance of infrared dim small targets, we propose a multi-domain attention-guided pyramid network (MAGPNet). Specifically, we design three modules to ensure that salient features of small targets can be acquired and retained in the multi-scale feature maps. To improve the adaptability of the network for targets of different sizes, we design a kernel aggregation attention block with a receptive field attention branch and weight the feature maps under different perceptual fields with attention mechanism. Based on the research on human vision system, we further propose an adaptive local contrast measure module to enhance the local features of infrared small targets. With this parameterized component, we can implement the information aggregation of multi-scale contrast saliency maps. Finally, to fully utilize the information within spatial and channel domains in feature maps of different scales, we propose the mixed spatial-channel attention-guided fusion module to achieve high-quality fusion effects while ensuring that the small target features can be preserved at deep layers. Experiments on public datasets demonstrate that our MAGPNet can achieve a better performance over other state-of-the-art methods in terms of the intersection of union, Precision, Recall, and F-measure. In addition, we conduct detailed ablation studies to verify the effectiveness of each component in our network. © 2024, Shanghai Jiao Tong University.

关键词： Software engineering

来源：评论

学校读者我要写书评

暂无评论

Towards Privacy-Aware Causal Structure Learning in Federated Setting

引用

IEEE Transactions on Big Data 2023年第6期9卷 1525-1535页

作者： Huang, Jianli Guo, Xianjie Yu, Kui Cao, Fuyuan Liang, Jiye Intelligent Interconnected Systems Laboratory of Anhui Province School of Computer Science and Information Engineering Hefei University of Technology Hefei230601 China Key Laboratory of Computational Intelligence and Chinese Information Processing of Ministry of Education School of Computer and Information Technology Shanxi University Taiyuan030006 China

Causal structure learning has been extensively studied and widely used in machine learning and various applications. To achieve an ideal performance, existing causal structure learning algorithms often need to centralize a large amount of data from multiple data sources. However, in the privacy-preserving setting, it is impossible to centralize data from all sources and put them together as a single dataset. To preserve data privacy, federated learning as a new learning paradigm has attached much attention in machine learning in recent years. In this paper, we study a privacy-aware causal structure learning problem in the federated setting and propose a novel federated PC (FedPC) algorithm with two new strategies for preserving data privacy without centralizing data. Specifically, we first propose a novel layer-wise aggregation strategy for a seamless adaptation of the PC algorithm into the federated learning paradigm for federated skeleton learning, then we design an effective strategy for learning consistent separation sets for federated edge orientation. The extensive experiments validate that FedPC is effective for causal structure learning in federated learning setting. © 2023 IEEE.

关键词： Big data

来源：评论

学校读者我要写书评

暂无评论

Deep-learning-based method for concealed object detection in terahertz (THz) images

Deep-learning-based method for concealed object detection in...

引用

2023 Advanced Fiber Laser Conference, AFL 2023

作者： Ge, Zihao Zhang, Yuan Wu, Xuyang Jia, Zhiyuan Wang, Heng Jia, Keke Key Laboratory of Grain Information Processing and Control [Henan University of Technology Ministry of Education Zhengzhou450001 China Henan Provincial Key Laboratory of Grain Information Detection and Control Zhengzhou450001 China College of Information Science and Engineering Henan University of Technology Zhengzhou450001 China

ISBN: (纸本)9781510677661

Terahertz (THz) technology has become a new trend in various fields due to its high penetration and harmlessness towards human body and objects. The object detection of concealed and hidden objects based on THz images is of great significance for ensuring public safety. However, the poor quality of original THz images leads to insufficient accuracy in target detection. Therefore, it is necessary to preprocess the images before performing object detection. In this work, in order to investigate the impact of different pre-processing methods on object detection using images, we adopt two methods, namely non-local mean (NLM) filtering and histogram equalization (HE). After pre-processing, YOLOv7 algorithm is used to perform object detection based on the preprocessed THz images. The experimental results show that YOLOv7 achieves highest recognition accuracy on NLM filtered THz images. The experimental results presented in this work provide a reference to select image processing method for performing concealed object detection based on THz images. © COPYRIGHT SPIE. Downloading of the abstract is permitted for personal use only.

关键词： Object detection

来源：评论

学校读者我要写书评

暂无评论

Lightweight Monocular 3D Object Detection Based on Attention Mechanism 5

Lightweight Monocular 3D Object Detection Based on Attention...

引用

5th Asia-Pacific Conference on Image processing, Electronics and Computers, IPEC 2024

作者： Zhan, Feng Zhang, Zhi Wuhan University of Science and Technology College of Computer Science and Technology Hubei Wuhan430065 China Information Processing and Real-time Industrial System Hubei Province Key Laboratory of Intelligent Hubei Wuhan430065 China Information Processing and Real-time Industrial System Hubei Province Key Laboratory of Intelligent Hubei Wuhan30065 China

ISBN: (纸本)9798350374407

In the field of autonomous driving, 3D target detection is an important technology. In view of the shortcomings of existing monocular 3D detection algorithms in terms of accuracy and real-time performance, we propose a lightweight 3D target detection algorithm based on the attention mechanism. In order to ensure the real-time performance of the model, we use depthwise separable convolutions in the backbone network. The feature maps of each scale are then connected to an EMA module, which can fuse spatial attention and cross-dimensional Coordinate attention is used to enhance the feature extraction capability of the network, and BiFPN is subsequently used for multi-scale feature fusion. Finally, experiments were conducted on the KITTI dataset, and the results showed that our algorithm was superior to other methods in both accuracy and speed. © 2024 IEEE.

关键词： Object recognition

来源：评论

学校读者我要写书评

暂无评论

intelligent Test Paper Generation Algorithm Integrating Normal Distribution and Parallel Genetic Algorithm 5

Intelligent Test Paper Generation Algorithm Integrating Norm...

引用

5th International Conference on intelligent Computing and Human-Computer Interaction, ICHCI 2024

作者： Liu, Chuangye Huang, Guimin School of Computer Science and Information Security Guilin University of Electronic Technology Guangxi Key Laboratory of Image and Graphic Intelligent Processing Guilin541004 China

ISBN: (纸本)9798350368284

Genetic algorithms have been widely used in intelligent test paper generation systems. However, traditional genetic algorithms cannot ensure that the difficulty of test questions is normally distributed, and are prone to falling into local optimal solutions. To address the above problems, we proposed an intelligent test paper generation algorithm that combines normal distribution and parallel genetic algorithms. In the population initialization stage, we use the cumulative distribution function (CDF) of the normal distribution to initialize the population. In the subsequent genetic operations, we designed a directed mutation mechanism and added KL divergence as a penalty number in the design of the fitness function to ensure that the difficulty of the test questions is normally distributed. At the same time, in order to speed up the efficiency of the algorithm and escape from the local optimal solution, an adaptive migration strategy was introduced. Experimental results show that the difficulty of the test questions generated by this algorithm is normally distributed, and the algorithm performance is better than other algorithms. © 2024 IEEE.

关键词： Consensus algorithm

来源：评论

学校读者我要写书评

暂无评论

A Dual-Branch Network Based on ViT and Mamba for Semantic Segmentation of Remote Sensing Image 2

A Dual-Branch Network Based on ViT and Mamba for Semantic Se...

引用

2nd IEEE International Conference on Signal, information and Data processing, ICSIDP 2024

作者： An, Ke Wang, Ying Chen, Liang Wang, Yupie Beijing Institute of Technology National Key Laboratory for Space-Born Intelligent Information Processing Beijing China Shanghai Academy of Spaceflight Technology Shanghai China

ISBN: (纸本)9798331515669

Semantic segmentation of remote sensing images has significant applications across various scenarios. The prevailing frameworks include Convolutional Neural Network (CNN) and Transformer. However, CNN is limited by the receptive field of convolutions, while the Transformer is constrained by computational complexity, which restricts attention calculations to local windows and fails to effectively address long-range dependency modeling. The efficient Mamba architecture, characterized by linear complexity, offers a promising solution to these challenges. Inspired by Mamba, we propose a dual-branch network based on ViT and Mamba. The Vision Transformer (ViT) branch incorporates the Swin Transformer to model spatial details while maintaining computational complexity within acceptable bounds. Complementarily, the Mamba branch efficiently captures global context and long-range dependencies. Additionally, to suppress noise and conflicting information arising from the fusion of features from different frameworks, we design the Cross-Model Fusion Module (CMFM) and the Cross-Model Relevance Loss (CMRLoss) to achieve semantic consistency in the fusion process. The comprehensive experimental findings on the commonly utilized GaoFen-2 and iSAID datasets clearly illustrate the advantages of our proposed approach compared to the leading methods in the field. © 2024 IEEE.

关键词： Semantic Segmentation

来源：评论

学校读者我要写书评

暂无评论

ResPocket: A Multi-Scale Feature Fusion Method for Improving Protein Binding Site Detection

ResPocket: A Multi-Scale Feature Fusion Method for Improving...

引用

2024 IEEE International Conference on Bioinformatics and Biomedicine, BIBM 2024

作者： Lin, Xiaoli Yang, Wenli Chen, Yaoyu Liao, Xiongwei Wu, Zimo Zhang, Xiaolong Wuhan University of Science and Technology Hubei Key Laboratory of Intelligent Information Processing and Real-Time Industrial System School of Computer Science and Technology China

ISBN: (纸本)9798350386226

Accurate detection of protein binding sites is critical for facilitating drug design. Most existing models rely on features at a single scale, leading to the loss of crucial global or local structural information. Therefore, this paper proposes a novel 3D model (ResPocket), to fully capture the multi-scale features of proteins. First, Residual Block (ResB) based on U-Net is used to capture features from different scales, including global and local structural information. Then, an Enhanced Attention Module (EAM) is designed to encode semantic dependencies by effectively capturing global context information. In addition, ResPocket focuses on learning local information about protein pockets with Self-Guided Mask Block (SGMB), to increase the robustness of the overall model and reduce its dependence on external mask guidance. Experimental results show that ResPocket improves by an average of 4.89% on DCA Top-n, by an average of 1.48% and 2.25% on DCC and DVO metrics, respectively. © 2024 IEEE.

关键词： Binding sites

来源：评论

学校读者我要写书评

暂无评论

CONVERGENCE ANALYSIS OF SOME FINITE ELEMENT PARALLEL ALGORITHMS FOR THE STATIONARY INCOMPRESSIBLE MHD EQUATIONS

引用

Journal of Computational Mathematics 2024年第1期42卷 49-70页

作者： Xiaojing Dong Yinnian He Hunan Key Laboratory for Computation and Simulation in Science and Engineering Key Laboratory of Intelligent Computing&Information Processing of Ministry of EducationSchool of Mathematics and Computational ScienceXiangtan UniversityXiangtan 411105China Institute of Applied Physics and Computational Mathematics Beijing 100088China School of Mathematics and Statistics Xi'an Jiaotong UniversityXi'an 710049China

By combination of iteration methods with the partition of unity method(PUM),some finite element parallel algorithms for the stationary incompressible magnetohydrodynamics(MHD)with different physical parameters are presented and *** algorithms are highly *** first,a global solution is obtained on a coarse grid for all approaches by one of the iteration *** parallelized residual schemes,local corrected solutions are calculated on finer meshes with overlapping *** subdomains can be achieved flexibly by a class of *** proposed algorithm is proved to be uniformly stable and ***,one numerical example is presented to confirm the theoretical findings.

关键词： Partition of unity method Local and parallel algorithm Finite element method Iteration methods Magnetohydrodynamics

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：