检索结果-内蒙古大学图书馆

arXiv 2024年

作者： Zhu, Yabin Wang, Qianwu Li, Chenglong Tang, Jin Huang, Zhixiang The Key Laboratory of Intelligent Computing and Signal Processing Ministry of Education Anhui University School of Public Safety and Emergency Management Anhui University of Science and Technology Hefei231131 China School of Artificial Intelligence Anhui University Hefei230601 China Information Materials and Intelligent Sensing Laboratory of Anhui Province Anhui Provincial Key Laboratory of Security Artificial Intelligence School of Artificial Intelligence Anhui University Hefei230601 China Information Materials and Intelligent Sensing Laboratory of Anhui Province Anhui Provincial Key Laboratory of Multimodal Cognitive Computation School of Computer Science and Technology Anhui University Hefei230601 China Center for Big Data and Population Health of IHM China

The complementary benefits from visible and thermal infrared data are widely utilized in various computer vision task, such as visual tracking, semantic segmentation and object detection, but rarely explored in Multiple Object Tracking (MOT). In this work, we contribute a large-scale Visible-Thermal video benchmark for MOT, called VT-MOT. VT-MOT has the following main advantages. 1) The data is large scale and high diversity. VT-MOT includes 582 video sequence pairs, 401k frame pairs from surveillance, drone, and handheld platforms. 2) The cross-modal alignment is highly accurate. We invite several professionals to perform both spatial and temporal alignment frame by frame. 3) The annotation is dense and high-quality. VT-MOT has 3.99 million annotation boxes annotated and double-checked by professionals, including heavy occlusion and object re-acquisition (object disappear and reappear) challenges. To provide a strong baseline, we design a simple yet effective tracking framework, which effectively fuses temporal information and complementary information of two modalities in a progressive manner, for robust visible-thermal MOT. A comprehensive experiment are conducted on VT-MOT and the results prove the superiority and effectiveness of the proposed method compared with state-of-the-art methods. From the evaluation results and analysis, we specify several potential future directions for visible-thermal MOT. The project is released in https://***/wqw123wqw/PFTrack. Copyright © 2024, The Authors. All rights reserved.

关键词： Semantic Segmentation

来源：评论

学校读者我要写书评

暂无评论

Concept lattice mining algorithm using rough entropy with variable precision thresholding and co-evolution

Jiefangjun Ligong Daxue Xuebao/Journal of PLA University of ...

引用

Jiefangjun Ligong Daxue Xuebao/Journal of PLA University of Science and technology (Natural Science Edition) 2011年第1期12卷 25-30页

作者： Ding, Wei-Ping Wang, Jian-Dong Guan, Zhi-Jin College of Computer Science and Technology Nanjing University of Aeronautics and Astronautics Nanjing 210016 China School of Computer Science and Technology Nantong University Nantong 226019 China Provincial Key Laboratory for Computer Information Processing Technology Soochow University Suzhou 215006 China

Based on some special advantages of variable precision rough sets model and co-evolutionary particle swarm algorithm, a novel concept lattice mining algorithm (REVPT)using rough entropy with variable precision thresholding and co-evolution was proposed to solve some optimization problems of the concept lattice mining. In this algorithm, variable precision rough entropy was used to scale the subpopulations of various concept lattices dynamically, and rough approximation lattices constructed. The global optimization efficiency of concept lattices was improved by sharing search experiences among different populations, which can reduce the scale of the former concept lattices, and deduce the consistent decision rule sets efficiently. The experimental results show that the proposed algorithm is better on the convergence and lattice optimization when the variable precision thresholding β is at a certain appropriate range. Therefore it is of robustness on the rules mining accuracy and efficiency.

关键词： Global optimization

来源：评论

学校读者我要写书评

暂无评论

Occlusion Scene Classification via Cascade Supervised Contrastive Learning

TechRxiv

引用

TechRxiv 2022年

作者： Xu, Jianming Li, Yunfei He, Li Li, Jun The Guangdong Provincial Key Laboratory of Urbanization and Geosimulation Center of Integrated Geographic Information Analysis School of Geography and Planning Sun Yatsen University Guangzhou510275 China The School of Computer Science China University of Geosciences Wuhan430078 China The Hubei Key Laboratory of Intelligent Geo-Information Processing China University of Geosciences Wuhan430078 China The School of Automation Science and Engineering South China University of Technology Guangzhou510640 China

Recent occlusion scene classification focuses on extracting occlusion-invariant features by considering class separability, unsupervised learning, and generalization ability, separately. Considering them together is profitable to address real-world occlusion problem. In this work, we propose cascade supervised contrastive learning to accommodate those considerations simultaneously. First, the pretext task uses rectangles to occlude the original images. Second, we use the siamese residual network to extract the features of original images, and pretext task augmented images. Third, compute the cascade supervised contrastive loss of the intermediate layers and final layer of the network. By optimizing the cascade supervised contrastive loss, the network can learn pretext-task invariant features. Besides, the pretext task-invariant features can adapt to the downstream occlusion-invariant feature extraction task due to the outstanding generalization ability of contrastive learning. Finally, a multilayer perceptron is used to perform scene classification on the extracted features. The proposed method is validated on two simulated datasets and tends to achieve performance improvement compared with typical unsupervised occlusion scene classification methods. The first application of contrastive learning for remote sensing image occlusion scene classification is the innovative contribution of this paper. This method can achieve accurate remote sensing scene classification with any degree and type of occlusion. © 2022, CC BY-NC-SA.

关键词： Remote sensing

来源：评论

学校读者我要写书评

暂无评论

STFormer: A dual-stage transformer model utilizing spatio-temporal graph embedding for multivariate time series forecasting

引用

Journal of Intelligent and Fuzzy Systems 2024年第3期46卷 6951-6967页

作者： Xiao, Yuteng Liu, Zhaoyang Yin, Hongsheng Wang, Xingang Zhang, Yudong Jinan China Jinan China Shandong Provincial Key Laboratory of Computer Networks Shandong Fundamental Research Center for Computer Science Jinan China Xuzhou University of Technology Xuzhou China School of Information and Control Engineering China University of Mining and Technology Xuzhou China School of Computing and Mathematical Sciences University of Leicester LeicesterLE1 7RH United Kingdom School of Computer Science and Engineering Southeast University Jiangsu Nanjing China

Multivariate Time Series (MTS) forecasting has gained significant importance in diverse domains. Although Recurrent Neural Network (RNN)-based approaches have made notable advancements in MTS forecasting, they do not effectively tackle the challenges posed by noise and unordered data. Drawing inspiration from advancing the Transformer model, we introduce a transformer-based method called STFormer to address this predicament. The STFormer utilizes a two-stage Transformer to capture spatio-temporal relationships and tackle the issue of noise. Furthermore, the MTS incorporates adaptive spatio-temporal graph structures to tackle the issue of unordered data specifically. The Transformer incorporates graph embedding to combine spatial position information with long-term temporal connections. Experimental results based on typical finance and environment datasets demonstrate that STFormer surpasses alternative baseline forecasting models and achieves state-of-the-art results for single-step horizon and multistep horizon forecasting. © 2024 - IOS Press. All rights reserved.

关键词： Recurrent neural networks

来源：评论

学校读者我要写书评

暂无评论

Expert-level vision-language foundation model for real-world radiology and comprehensive evaluation

arXiv

引用

arXiv 2024年

作者： Liu, Xiaohong Yang, Guoxing Luo, Yulin Mao, Jiaji Zhang, Xiang Gao, Ming Zhang, Shanghang Shen, Jun Wang, Guangyu Department of Radiology Sun Yat-Sen Memorial Hospital Guangdong Provincial Key Laboratory of Malignant Tumor Epigenetics and Gene Regulation Medical Research Center Sun Yat-Sen Memorial Hospital Sun Yat-Sen University Guangzhou China State Key Laboratory of Networking and Switching Technology Beijing University of Posts and Telecommunications Beijing China National Key Laboratory for Multimedia Information Processing School of Computer Science Peking University Beijing China

Radiology is a vital and complex component of modern clinical workflow and covers many tasks. Recently, vision-language (VL) foundation models in medicine have shown potential in processing multimodal information, offering a unified solution for various radiology tasks. However, existing studies either pre-trained VL models on natural data or did not fully integrate vision-language architecture and pretraining, often neglecting the unique multimodal complexity in radiology images and their textual contexts. Additionally, their practical applicability in real-world scenarios remains underexplored. Here, we present RadFound, a large and open-source vision-language foundation model tailored for radiology, that is trained on the most extensive dataset of over 8.1 million images and 250,000 image-text pairs, covering 19 major organ systems and 10 imaging modalities. To establish expert-level multimodal perception and generation capabilities, RadFound introduces an enhanced vision encoder to capture intra-image local features and inter-image contextual information, and a unified cross-modal learning design tailored to radiology. To fully assess the models' capability, we construct a benchmark, RadVLBench, including radiology interpretation tasks like medical vision-language question-answering, as well as text generation tasks ranging from captioning to report generation. We also propose a human evaluation framework. When evaluated on the real-world benchmark involving three representative modalities, 2D images (chest X-rays), multi-view images (mammograms), and 3D images (thyroid CT scans), RadFound significantly outperforms other VL foundation models on both quantitative metrics and human evaluation. In summary, the development of RadFound represents an advancement in radiology generalists, demonstrating broad applicability potential for integration into clinical workflows. © 2024, CC BY.

关键词： Visual languages

来源：评论

学校读者我要写书评

暂无评论

Unsupervised Time Series Anomaly Detection under Data Contamination

Unsupervised Time Series Anomaly Detection under Data Contam...

引用

IEEE International Symposium on information (IT) in Medicine and Education, ITME

作者： Xiaohui Lin Zuoyong Li Xunhua Huang Xinwei Chen Haoyi Fan College of Computer and Data Science Fuzhou University Fuzhou China Fujian Provincial Key Laboratory of Information Processing and Intelligent Control College of Computer and Control Engineering Minjiang University Fuzhou China School of Computer Science and Technology Harbin University of Science and Technology Harbin China Fujian Engineering and Research Center of New Chinese Lacquer Materials Minjiang University Fuzhou China School of Computer and Artificial Intelligence Zhengzhou University Zhengzhou China

Unsupervised learning utilizes unlabeled data to alleviate the reliance on large amounts of labeled data, and it has made great progress in time series anomaly detection. However, there are still some thorny issues unsolved, especially the incapability to maintain the robustness of the model when the time series dataset is contaminated. In this work, we propose an unsupervised time series anomaly detection method on data contamination, named distribution alignment autoencoder (DAAE), in which a temporal sliding reward function combines the adversarial distribution alignment to mine the contaminated data. Specifically, the dimensionality reduction and reconstruction of the time series are implemented by an autoencoder. Next, using two discriminators, the latent features obtained by dimensionality reduction are mapped to the Gaussian distribution, and the high-confidence normal time series are generated by sampling noise from the Gaussian distribution. Then, we compute the mean square error between high-confidence normal samples and the training data. Finally, we feed the error values to the temporal sliding reward function to mine the contaminated training samples. Experimental results compared with baselines on public datasets demonstrate the effectiveness of DAAE.

关键词： Training Dimensionality reduction Time series analysis Training data Mean square error methods Gaussian distribution Robustness

来源：评论

学校读者我要写书评

暂无评论

Efficient Gridless DoA Estimation Method of Non-uniform Linear Arrays with Applications in Automotive Radars

arXiv

引用

arXiv 2023年

作者： Gao, Silin Zhang, Zhe Wang, Muhan Zhang, Yan Zhao, Jie Zhang, Bingchen Wang, Yue Wu, Yirong Aerospace Information Research Institute Chinese Academy of Sciences Beijing China Key Laboratory of Technology in Geo-spatial Information Processing and Application System Chinese Academy of Sciences Beijing China School of Electronic Electrical and Communication Engineering University of Chinese Academy of Sciences Beijing China Suzhou Key Laboratory of Microwave Imaging Processing and Application Technology Suzhou Aerospace Information Research Institute Jiangsu Suzhou China The Beijing Autoroad Tech Co. Ltd China Electrical and Computer Engineering Department George Mason University FairfaxVA United States

This paper focuses on the gridless direction-of-arrival (DoA) estimation for data acquired by non-uniform linear arrays (NLAs) in automotive applications. Atomic norm minimization (ANM) is a promising gridless sparse recovery algorithm under the Toeplitz model and solved by convex relaxation, thus it is only applicable to uniform linear arrays (ULAs) with array manifolds having a Vandermonde structure. In automotive applications, it is essential to apply the gridless DoA estimation to NLAs with arbitrary geometry with efficiency. In this paper, a fast ANM-based gridless DoA estimation algorithm for NLAs is proposed, which employs the array manifold separation technique and the accelerated proximal gradient (APG) technique, making it applicable to NLAs without losing of efficiency. Simulation and measurement experiments on automotive multiple-input multiple-outpt (MIMO) radars demonstrate the superiority of the proposed method. Copyright © 2023, The Authors. All rights reserved.

关键词： Efficiency

来源：评论

学校读者我要写书评

暂无评论

Asymmetry Total Variation and Framelet Regularized Nonconvex Low-Rank Tensor Completion

SSRN

引用

SSRN 2022年

作者： Chen, Yongyong Zhao, Xiaojia Xu, Tingting Zeng, Haijin Xu, Yanhui Chen, Junxing The School of Computer Science and Technology Harbin Institute of Technology Shenzhen518055 China Guangdong Provincial Key Laboratory of Novel Security Intelligence Technologies Shenzhen518055 China Image Processing and Interpretation Imec Research Group Ghent University Belgium Department of Electrical and Computer Engineering Faculty of Science and Technology University of Macau China The College of Medicine and Biological Information Engineering Northeastern University Shenyang110004 China

The low-rank tensor representation has shown enormous potential and advantages in diverse applications. However, (1) due to the inherent limitations of the low-rank tensor model, the tensor nuclear norm is utilized yet as a biased estimate of the tensor rank, which limits the image recovery performance. (2) most current low-rank approximation-based methods focus on the preservation of global information, while ignoring the local details preservation such as the spatial piece-wise smoothness and sharp edges. To overcome he above obstacles, we propose a novel asymmetry three-dimensional total variation and framelet regularized nonconvex low-rank tensor completion (ATV-FNTC) model, which integrates the nonconvex penalty function and asymmetric three-dimensional total variation into one unified model. Specifically, ATV-FNTC introduces one nonconvex penalty function as a tighter regularizer to approximate the tensor rank. Different from existing methods, the asymmetric three-dimensional total variation regularization is developed to achieve more accurate recovery of detailed information and flexibly control the smoothing strength of different dimensions of tensor data. Furthermore, we design an iterative algorithm to solve the nonconvex ATV-FNTC model by the alternating direction methods of multipliers. Experimental results reveal the superiority of the proposed method compared with several state-of-the-art methods when handling different completion tasks. © 2022, The Authors. All rights reserved.

关键词： Tensors

来源：评论

学校读者我要写书评

暂无评论

Optimizing the Taxonomy and Hierarchy of a Chinese Lexical Database - Cilin

Optimizing the Taxonomy and Hierarchy of a Chinese Lexical D...

引用

International Conference on Behavior, Economic and Social Computing (BESC)

作者： Tingxin Wei Xiaohe Chen Shi Shengwang Junsheng Zhou Yanhui Gu Weiguang Qu International College for Chinese Studies Nanjing Normal University Nanjing China School of Chinese Language and Literature Nanjing Normal University Nanjing China School of Computer Science and Technology Nanjing Normal University Nanjing China Fujian Provincial Key Laboratory of Information Processing and Intelligent Control Minjiang University Fuzhou China

ISBN: (纸本)9781728102085;9781728102078

Cilin is one of the most popular semantic knowledge bases in Chinese information processing. Due to its coding scheme and taxonomical arrangement, some semantic relations among words are not explicitly shown. Our work aims to characterize its semantic relations by adding tags and compound codes to optimize the taxonomy and hierarchy of Cilin. Experiment results show that using the tag-augmented Cilin as a knowledge base improves the performance in the task of semantic similarity measurement.

关键词： Semantics Compounds Taxonomy Encoding Knowledge based systems Wheels Task analysis

来源：评论

学校读者我要写书评

暂无评论

Syndrome Classification Based on Multi-Graph Attention Network

Syndrome Classification Based on Multi-Graph Attention Netwo...

引用

IEEE International Symposium on information (IT) in Medicine and Education, ITME

作者： Jishun Ma Weikai Lu Changen Zhou Shenghua Teng Zuoyong Li College of Electronic and Information Engineering Shandong University of Science and Technology Qingdao China School of Electronic Electrical Engineering and Physics Fujian University of Technology Fuzhou China Research Base of Traditional Chinese Medicine Syndrome Fujian University of Traditional Chinese Medicine Fuzhou China Fujian Provincial Key Laboratory of Information Processing and Intelligent Control College of Computer and Control Engineering Minjiang University Fuzhou China

Syndrome classification is an important step in Traditional Chinese Medicine (TCM) for diagnosis and treatment. In this paper, we propose a multi-graph attention network (MGAT) based method to simulate TCM doctors to infer the syndromes. Specifically, the complex relationships between symptoms and state elements are aggregated using graph attention networks (GAT) and syndromes are classified by a multilayer perceptron (MLP). To verify the effectiveness of the model, extensive experiments are conducted on the Treatise on Febrile Diseases dataset. The experimental results show that the proposed method outperforms several typical methods in terms of accuracy, precision, recall, and F1-score. The MGAT model has high accuracy in syndrome classification and has practical application value.

关键词： Computational modeling Education Medical services Multilayer perceptrons Feature extraction computer aided diagnosis information technology

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：