检索结果-内蒙古大学图书馆

Learning Video Salient Object Detection Progressively From Unlabeled Videos

IEEE Transactions on Multimedia 2025年 27卷 2423-2435页

作者： Xu, Binwei Jiang, Qiuping Liang, Haoran Zhang, Dingwen Liang, Ronghua Chen, Peng Ningbo University Faculty of Information Science and Engineering Ningbo315211 China Zhejiang University of Technology College of Computer Science and Technology Hangzhou310023 China Northwestern Polytechnical University Brain and Artificial Intelligence Laboratory School of Automation Xi'an710072 China

Recently, deep learning-based video salient object detection (VSOD) has achieved some breakthroughs, but these methods rely on expensive annotated videos with pixel-wise annotations or weak annotations. In this paper, based on the similarities and differences between VSOD and image salient object detection (SOD), we propose a novel VSOD method via a progressive framework that locates and segments salient objects in sequence without utilizing any video annotation. To efficiently use the knowledge learned in the SOD dataset for VSOD efficiently, we introduce dynamic saliency to compensate for the lack of motion information of SOD during the locating process while maintaining the same fine segmenting process. Specifically, we utilize the coarse locating model trained on the image dataset, to identify frames with both static and dynamic saliency. Locating results of these frames are selected as spatiotemporal location labels. Moreover, by tracking salient objects in adjacent frames, the number of spatiotemporal location labels is increased. On the basis of these location labels, a two-stream locating network with an optical flow branch is proposed to capture salient objects in videos. The results with respect to five public benchmarks demonstrate that our method outperforms the state-of-the-art weakly and unsupervised methods. © 1999-2012 IEEE.

关键词： Self-supervised learning

来源：评论

学校读者我要写书评

暂无评论

EM Algorithm for Cluster of Multicore Nodes Using Level 3 BLAS Operations to Learn Gaussian Mixture Models 15th

EM Algorithm for Cluster of Multicore Nodes Using Level 3 ...

引用

15th International Conference on Parallel Processing and Applied Mathematics, PPAM 2024

作者： Kwedlo, Wojciech Faculty of Computer Science Bialystok University of Technology Wiejska 45A Bialystok15-351 Poland

ISBN: (纸本)9783031857027

In this paper, we explore the problem of learning the parameters of Gaussian mixture models using the expectation-maximization (EM) algorithm. We propose a new parallel formulation of the EM algorithm that utilizes a static decomposition and distributes the learning set and the matrix storing posterior probabilities among MPI processes. During both the E-step and M-step of an EM iteration, each MPI process spawns a team of OpenMP threads that process data in blocks sized to fit the last-level cache. The calculations necessary for obtaining weighted Gaussian densities in the E-step and mixture parameters in the M-step are conducted using optimized level 3 BLAS operations. Hierarchical all-reduce operations are employed to compute the sums required in the M-step, initially among the OpenMP threads within each MPI process and subsequently among all the MPI processes in the parallel application. In our computational experiments, we compared this proposed approach with the traditional method that employs level 2 BLAS operations and assessed its strong scaling on 64 nodes of a compute cluster. The results demonstrate that the proposed method is 1.6 to 4.7 times faster than the conventional approach utilizing level 2 BLAS. The parallel efficiency of our approach on 64 nodes ranges from 60% to 73%. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Expectation maximization algorithm

来源：评论

学校读者我要写书评

暂无评论

A novel perturbation-based degraded image super-resolution method for object recognition in intelligent transportation system

引用

Neural Computing and Applications 2025年第6期37卷 5169-5186页

作者： Zhang, Cheng Zeng, Shan Yang, Zhiguang Chen, Yulong Li, Hao Tang, Yuanyan College of Mathematics and Computer Science Wuhan Polytechnic University Hubei Wuhan430023 China College of Medicine and Health Science Wuhan Polytechnic University Hubei Wuhan430023 China Faculty of Science and Technology University of Macau 999078 China

Intelligent transportation system (ITS) plays an important role in assisting drivers to master road information and optimize traffic flow. However, image degradation resulting from complicated environmental factors, such as motion blur caused by vehicle movement and illumination condition, caused some difficulties in current object recognition research on the ITS that may pose serious risks to driving safety. In order to tackle these challenges, this paper introduces a novel perturbation-based image super-resolution method based on GAN inversion (PSRGANI), utilizing a perturbation mechanism to better assist the latent space escape from the local optimum. In the architecture of PSRGANI, a dual encoder that preserves high-dimensional semantic feature from degraded images, extracts texture information from high-resolution (HR) images and completes the SR reconstruction process via perturbation mechanism. The additional encoder effectively decreases illuminated interference from external environment, enhancing the result robustness of SR reconstruction. A dual discriminator precisely regulates the upsampling process of the decoder for improving the quality of generated images. The additional discriminator achieves the optimal fusion of inputs from different sources by decoupling. SR experiment results reveal the higher evaluation metrics of PSRGANI in texture and decoupling compared with other SR models such as ESRGAN and DGP. In real-world ITS of traffic sign experiments, PSRGANI-applied model shows better performance (Top-1 Class Error of 3%) and faster inference speed (0.06 s per image) when compared to target detection algorithms such as YOLO and DETR. PSRGANI is demonstrated to have accurate results on degraded image recognition in terms of texture quality and evaluation metrics. © The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature 2024.

关键词： Image reconstruction

来源：评论

学校读者我要写书评

暂无评论

EFS-XGBoost: A robust framework for precision classification of COVID-19 cases

引用

Intelligent Data Analysis 2025年第1期29卷 171-185页

作者： Abidi, Mustufa Haider Khare, Neelu D, Preethi Alkhalefah, Hisham Umer, Usama Advanced Manufacturing Institute King Saud University Riyadh Saudi Arabia School of Computer Science Engineering and Information Systems Vellore Institute of Technology Tamil Nadu Vellore India Department of Computer Science and Engineering Faculty of Engineering and Technology SRM Institute of Science and Technology Ramapuram Chennai India

The emergence of the novel COVID-19 virus has had a profound impact on global healthcare systems and economies, underscoring the imperative need for the development of precise and expeditious diagnostic tools. Machine learning techniques have emerged as a promising avenue for augmenting the capabilities of medical professionals in disease diagnosis and classification. In this research, the EFS-XGBoost classifier model, a robust approach for the classification of patients afflicted with COVID-19 is proposed. The key innovation in the proposed model lies in the Ensemble-based Feature Selection (EFS) strategy, which enables the judicious selection of relevant features from the expansive COVID-19 dataset. Subsequently, the power of the eXtreme Gradient Boosting (XGBoost) classifier to make precise distinctions among COVID-19-infected patients is *** EFS methodology amalgamates five distinctive feature selection techniques, encompassing correlation-based, chi-squared, information gain, symmetric uncertainty-based, and gain ratio approaches. To evaluate the effectiveness of the model, comprehensive experiments were conducted using a COVID-19 dataset procured from Kaggle, and the implementation was executed using Python programming. The performance of the proposed EFS-XGBoost model was gauged by employing well-established metrics that measure classification accuracy, including accuracy, precision, recall, and the F1-Score. Furthermore, an in-depth comparative analysis was conducted by considering the performance of the XGBoost classifier under various scenarios: employing all features within the dataset without any feature selection technique, and utilizing each feature selection technique in isolation. The meticulous evaluation reveals that the proposed EFS-XGBoost model excels in performance, achieving an astounding accuracy rate of 99.8%, surpassing the efficacy of other prevailing feature selection techniques. This research not only advances the field of COVI

关键词： classification COVID-19 ensemble-based feature selection machine learning XGBoost

来源：评论

学校读者我要写书评

暂无评论

Latency- and Privacy-Aware Convolutional Neural Network Distributed Inference for Reliable Artificial Intelligence Systems

IEEE Transactions on Artificial Intelligence

引用

IEEE Transactions on Artificial Intelligence 2025年第2期6卷 365-377页

作者： Hu, Yuhao Xu, Xiaolong Qi, Lianyong Zhou, Xiaokang Xia, Xiaoyu Nanjing University of Information Science and Technology School of Software Nanjing210044 China College of computer science and technology Qingdao266850 China Shiga University Faculty of Data Science Hikone522-8522 Japan RMIT University School of Computing Technologies MelbourneVIC3000 Australia

Reliable artificial intelligence (AI) systems not only propose a challenge on providing intelligent services with high quality for customers but also require customers' privacy to be protected as much as possible during process of the services. Given the ultrahigh computing load brought by deep-learning-based intelligent services to edge devices and the ultralong distance between edge and cloud, low-latency requirement of intelligent services is hard to meet in single edge computing or cloud computing. Edge-cloud collaborative inference of deep neural networks (DNNs) is considered a feasible solution to the problem. However, former work has not reduced the inference latency to the greatest extent and has not considered privacy protection in distributed systems. To solve the problem, we first establish a novel queue mechanism. Then, the convolution layer split decisions are made based on deep reinforcement learning (DRL) to realize the parallel inference of convolutional neural networks (CNNs) for inference latency reduction. Next, for each CNN, the partition decision is made based on brute force algorithm to further reduce inference latency and protect customers' privacy. Finally, simulation results show that our method performs better than the existing other methods. © 2020 IEEE.

关键词： Cloud computing

来源：评论

学校读者我要写书评

暂无评论

Superconductivity in Bi1.6Pb0.4Sr2Ca2Cu3O10+δ ceramics with cobalt ferrite (CoFe2O4) nanoparticle addition

引用

Ceramics International 2025年第12期51卷 16134-16142页

作者： Tran, Duc H. Pham, An T. Binh, Nguyen Thanh Anh, Do Thi Kim Nam, Nguyen Hoang Hong, Nguyen T.M. Cuong, Le Viet Pham, Phong V. Huy, Nguyen Quang Huong, Tran Thi Man, Nguyen K. Lim, Kean Pah Kechik, Mohd Mustafa Awang Chen, Soo Kien Le, Tien Faculty of Physics University of Science Vietnam National University Hanoi100000 Viet Nam Nano and Energy Center University of Science Vietnam National University Hanoi100000 Viet Nam Faculty of Engineering Physics and Nanotechnology VNU University of Engineering and Technology Hanoi100000 Viet Nam University of Science Vietnam National University Hanoi100000 Viet Nam Faculty of Electronics and Computer Engineering University of Economics - Technology for Industries Hanoi100000 Viet Nam Faculty of Electronic Materials and Devices School of Materials Science and Engineering Hanoi University of Science and Technology Hanoi100000 Viet Nam Department of Physics Faculty of Science Universiti Putra Malaysia Selangor Serdang43400 Malaysia University of Science and Technology of Hanoi Vietnam Academy of Science and Technology 18 Hoang Quoc Viet Cau Giay Hanoi100000 Viet Nam

This study explores the use of cobalt ferrite (CoFe2O4) nanoparticles as artificial pinning centers to investigate flux pinning characteristics in the Bi1.6Pb0.4Sr2Ca2Cu3O10+δ ceramics. Polycrystalline samples with the composition (Bi1.6Pb0.4Sr2Ca2Cu3O10+δ)1−x(CoFe2O4)x, where x ranges from 0.000 to 0.010, are synthesized based on the solid-state formation. Structural and surface analyses reveal that the samples predominantly comprise superconducting phases, with volume fractions varying with x. An increase in x leads to a drop in the critical temperature (Tc). However, an improvement in the critical current density (Jc) is observed at x = 0.002, owing to the addition of CoFe2O4. The collective pinning theory indicates a shift from a single-vortex to small bundle vortex pinning at the crossover field (Bsb) in all samples. The dependence of Bsb on temperatures is consistent with δl pinning, attributed to changes in the mean free path of the charge carrier. Geometries of artificial pinning centers are identified as 2D using the Dew–Hughes model. CoFe2O4 nanoparticles are primarily located at the grain boundaries, filling voids and supporting pinning through grain boundaries. © 2025 Elsevier Ltd and Techna Group S.r.l.

关键词： Flux pinning

来源：评论

学校读者我要写书评

暂无评论

REHG: A Recommender Engine Based on Heterogeneous Graph 18th

REHG: A Recommender Engine Based on Heterogeneous Graph

引用

18th International Conference on Wireless Artificial Intelligent Computing Systems and Applications, WASA 2024

作者： Xin, Xiaoyang Cheng, Zesheng Wang, Tiankuan Yan, Mengqiu Zhao, Ruixuan Peng, Chong Li, Jianbo College of Computer Science and Technology Qingdao University Qingdao266071 China Faculty of Computing Science University of Alberta Edmonton Canada

ISBN: (纸本)9783031714634

Recommendation systems are extensively utilized across diverse domains with the aim of suggesting the most appropriate items to individual users. An effective recommendation engine provides valuable guidance for daily life. Therefore, improving the efficiency of recommender systems has become a prominent subject of interest in both industrial and academic circles. The majority of high-performing models in recommendation tasks utilize binary homogeneous graph convolutional methods and collaborative filtering as their foundational structures. These models are simple in structure, accurate and efficient in recommendation tasks. However, owing to the constraints of the binary homogeneous graph's structure, these models are unable to incorporate supplementary information during operation, thereby hindering the full utilization of diverse and efficient data. To fill the research gaps, this study introduces a novel recommendation system named REHG, which is grounded on directed heterogeneous graphs. Additionally, the study devises two experiments to assess its performance across various testing datasets. These experiments offer compelling evidence to support the superior performance of the proposed REHG model over common baseline models across different evaluation criteria. Furthermore, the study explores the feasibility of enhancing interoperability among various recommender system models. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Collaborative filtering

来源：评论

学校读者我要写书评

暂无评论

Integration of Convolutional Neural Network and Image Processing for Pulp Fibril Detection and Measurement

引用

IEEE Access 2025年 13卷 74634-74646页

作者： Chirakitsakul, Tanachot Wattuya, Pakaket Somboon, Phichit Jansakra, Panthira Watcharopas, Chakrit Kasetsart University Department of Computer Science Bangkok Thailand Kasetsart University Artificial Intelligence Technology and Innovation Center for Health Faculty of Science Bangkok Thailand Kasetsart University Department of Forest Products Bangkok Thailand

The fibrillation index is a critical metric in paper manufacturing, quantifying the degree of fibrillation achieved during the pulp refining process. Optimizing this metric enhances both paper quality and production efficiency. However, traditional measurement methods—such as manual visual examination of pulp samples under microscopy—are slow, error-prone, and labor-intensive, limiting their scalability in industrial applications. This study proposes a novel method that integrates deep learning with image processing techniques to automate fibril detection and fibrillation index computation. The proposed method leverages the discriminative capabilities of convolutional neural networks (CNNs) with adaptive image processing techniques to overcome key challenges such as low contrast, image noise, and variability in fibril morphology. The patch-based classification approach effectively filters out irrelevant objects, especially those whose features visually resemble fibrils, thus improving fibril segmentation accuracy. The method was comprehensively validated against expert-labeled ground truth images and achieved a promising average error rate of 0.4494 ± 0.4187. Experimental results also demonstrate the strong robustness of the proposed method, with consistent performance across diverse refining conditions and image qualities, making it suitable for real-world application in the pulp and paper industry. Furthermore, this study paves the way for broader applications in materials science and biomedical imaging, where precise feature detection in microscopic images is essential. © 2013 IEEE.

关键词： Pulp refining

来源：评论

学校读者我要写书评

暂无评论

Improved GPU Memory Management in Evolutionary Decision Tree Induction for Large-Scale Data 15th

Improved GPU Memory Management in Evolutionary Decision Tre...

引用

15th International Conference on Parallel Processing and Applied Mathematics, PPAM 2024

作者： Jurczuk, Krzysztof Reska, Daniel Czajkowski, Marcin Kretowski, Marek Faculty of Computer Science Bialystok University of Technology Wiejska 45a Bialystok15-351 Poland

ISBN: (纸本)9783031856969

Decision trees (DTs) are explainable machine learning techniques applicable to classification and regression problems. Traditionally, DTs are built through a top-down greedy search, which is usually fast but may lead to sub-optimal solutions. An alternative approach involves the use of evolutionary algorithms (EAs), which allow for more global exploration that can yield simpler and accurate DTs. However, the EA-based DT induction is computationally demanding, especially for large-scale data. To alleviate these high computing requirements, various parallel and distributed accelerations are continuously explored. In this paper, we focus on a GPU-supported solution and extend it by improving memory management. To reduce synchronization points between GPU threads (both within and across blocks), the new solution avoids atomic functions. However, such an approach requires allocating individual result buffers for each thread, which increases memory requirements. To compensate for this, a compact in-memory representation of DTs is additionally applied. Moreover, an additional level of reduction is necessary. Experimental validation on various datasets, both artificial and real-life, shows that the enhanced solution further accelerates the EA-based DT induction. The results also reveal that the time savings increase as the dataset size grows and are influenced by DT size. Therefore, the hybrid solution provides the best time results, by applying synchronization avoidance for smaller DTs and the opposite strategy for larger ones. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Memory management

来源：评论

学校读者我要写书评

暂无评论

RobustFace: a novel image restoration technique for face adversarial robustness improvement

引用

Multimedia Tools and Applications 2025年第11期84卷 8965-8993页

作者： Sadu, Chiranjeevi Das, Pradip K. Yannam, V Ramanjaneyulu Nayyar, Anand Department of Computer Science and Engineering RGUKT Nuzvid Andhra Pradesh Nuzvid India Department of Computer Science and Engineering IIT Guwahati Assam Guwahati India School of Technology Woxsen University Telangana Hyderabad India Graduate School Faculty of Information Technology Duy Tan University Da Nang550000 Viet Nam

Machine Learning (ML) models, particularly Deep Learning (DL), have made rapid progress and achieved significant milestones across various applications, including numerous safety-critical contexts. However, these models have recently been discovered to be susceptible to adversarial attacks, which are well-crafted input images. The primary objective of this paper is to propose a novel methodology titled "RobustFace", which is based on deep image restoration networks, that significantly improves the facial adversarial robustness of various image-classification models. Adversarial images are created using the Private Fast Gradient Sign Method (P-FGSM), StyleGAN and Fast Landmark Manipulation (FLM) methods. The adversarial images are then enhanced using deep image restoration networks to bring back them into the original space. The encoded weighted local magnitude patterns (WLMP) are extracted and provided to different types of classifiers to detect facial adversarial images from the clean images. The effectiveness of RobustFace has been demonstrated on two real-world datasets and experimental outcomes show that it significantly improves facial adversarial robustness on all evaluating classifiers. It improves the highest classification accuracy from 98.75% to 99.00% on P-FGSM attacks, from 77.94% to 85.25% on adversarial attacks generated by StyleGAN and from 65.52% to 69.50% for FLM attacks. © The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer Nature 2024.

关键词： Image reconstruction

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：