检索结果-内蒙古大学图书馆

arXiv 2022年

作者： Su, Yukun Deng, Jingliang Sun, Ruizhou Lin, Guosheng Wu, Qingyao The School of Software Engineering South China University of Technology Guangzhou510640 China The School of Computer Science and Engineering Nanyang Technological University Singapore Singapore

Humans tend to mine objects by learning from a group of images or several frames of video since we live in a dynamic world. In the computer vision area, many researches focus on co-segmentation (CoS), co-saliency detection (CoSD) and video salient object detection (VSOD) to discover the co-occurrent objects. However, previous approaches design different networks on these similar tasks separately, and they are difficult to apply to each other, which lowers the upper bound of the transferability of deep learning frameworks. Besides, they fail to take full advantage of the cues among inter- and intra-feature within a group of images. In this paper, we introduce a unified framework to tackle these issues, term as UFO (Unified Framework for Co-Object Segmentation). Specifically, we first introduce a transformer block, which views the image feature as a patch token and then captures their long-range dependencies through the self-attention mechanism. This can help the network to excavate the patch structured similarities among the relevant objects. Furthermore, we propose an intra-MLP learning module to produce self-mask to enhance the network to avoid partial activation. Extensive experiments on four CoS benchmarks (PASCAL, iCoseg, Internet and MSRC), three CoSD benchmarks (Cosal2015, CoSOD3k, and CocA) and four VSOD benchmarks (DAVIS16, FBMS, ViSal and SegV2) show that our method outperforms other state-of-the-arts on three different tasks in both accuracy and speed by using the same network architecture, which can reach 140 FPS in real-time. Code is available at https://***/suyukun666/UFO © 2022, CC0.

关键词： Chemical activation

来源：评论

学校读者我要写书评

暂无评论

Cactus Mechanisms: Optimal Differential Privacy Mechanisms in the Large-Composition Regime

Cactus Mechanisms: Optimal Differential Privacy Mechanisms i...

引用

IEEE International Symposium on Information Theory

作者： Wael Alghamdi Shahab Asoodeh Flavio P. Calmon Oliver Kosut Lalitha Sankar Fei Wei School of Engineering and Applied Science Harvard University Department of Computing and Software McMaster University School of Electrical Computer and Energy Engineering Arizona State University

ISBN: (数字)9781665421591

ISBN: (纸本)9781665421607

Most differential privacy mechanisms are applied (i.e., composed) numerous times on sensitive data. We study the design of optimal differential privacy mechanisms in the limit of a large number of compositions. As a consequence of the law of large numbers, in this regime the best privacy mechanism is the one that minimizes the Kullback-Leibler divergence between the conditional output distributions of the mechanism given two different inputs. We formulate an optimization problem to minimize this divergence subject to a cost constraint on the noise. We first prove that additive mechanisms are optimal. Since the optimization problem is infinite dimensional, it cannot be solved directly; nevertheless, we quantize the problem to derive nearoptimal additive mechanisms that we call "cactus mechanisms" due to their shape. We show that our quantization approach can be arbitrarily close to an optimal mechanism. Surprisingly, for quadratic cost, the Gaussian mechanism is strictly suboptimal compared to this cactus mechanism. Finally, we provide numerical results which indicate that cactus mechanisms outperform Gaussian and Laplace mechanisms for a finite number of *** full proofs can be found in the extended version at [1]. This paper is Part I in a pair of papers, where Part II is [2].

关键词： Differential privacy Privacy Costs Additives Quantization (signal) Shape Optimization

来源：评论

学校读者我要写书评

暂无评论

Idea23D: Collaborative LMM Agents Enable 3D Model Generation from Interleaved Multimodal Inputs

arXiv

引用

arXiv 2024年

作者： Chen, Junhao Li, Xiang Ye, Xiaojun Li, Chao Fan, Zhaoxin Zhao, Hao Tsinghua University China Tsinghua Shenzhen International Graduate School Tsinghua University China School of Software and Microelectronics Peking University China College of Computer Science Zhejiang University China College of Computer Science and Technology Harbin Engineering University China Beijing Advanced Innovation Center for Future Blockchain and Privacy Computing School of Artificial Intelligence Beihang University China

With the success of 2D diffusion models, 2D AIGC content has already transformed our lives. Recently, this success has been extended to 3D AIGC, with state-of-the-art methods generating textured 3D models from single images or text. However, we argue that current 3D AIGC methods still don’t fully unleash human creativity. We often imagine 3D content made from multimodal inputs, such as what it would look like if my pet bunny were eating a doughnut on the table. In this paper, we explore a novel 3D AIGC approach: generating 3D content from IDEAs. An IDEA is a multimodal input composed of text, image, and 3D models. To our knowledge, this challenging and exciting 3D AIGC setting has not been studied before. We propose the new framework Idea23D, which combines three agents based on large multimodal models (LMMs) and existing algorithmic tools. These three LMM-based agents are tasked with prompt generation, model selection, and feedback reflection. They collaborate and critique each other in a fully automated loop, without human intervention. The framework then generates a text prompt to create 3D models that align closely with the input IDEAs. We demonstrate impressive 3D AIGC results that surpass previous methods. To comprehensively assess the 3D AIGC capabilities of Idea23D, we introduce the Eval3DAIGC-198 dataset, containing 198 multimodal inputs for 3D generation tasks. This dataset evaluates the alignment between generated 3D content and input IDEAs. Our user study and quantitative results show that Idea23D significantly improves the success rate and accuracy of 3D generation, with excellent compatibility across various LMM, Text-to-Image, and Image-to-3D models. Code and dataset are available at https://***/. © 2024, CC BY-NC-SA.

关键词： 3D modeling

来源：评论

学校读者我要写书评

暂无评论

A Constrained Decomposition Grids Approach for U-shaped Disassembly Line Balancing Problem Considering Destructive Disassembly

A Constrained Decomposition Grids Approach for U-shaped Disa...

引用

Chinese Control and Decision Conference, CCDC

作者： Wenzheng Chen Shujin Qin XiWang Guo Jiacun Wang Qi Zhang SiQi Lei School of Economics and Management Shangqiu Normal University Shangqiu China Liaoning Petrochemical University Monmouth University West Long Branch USA Computer Science and Software Engineering Monmouth University West Long Branch USA Information Science and Engineering Shenyang University of Chemical Technology Shenyang China Computer and Communication Engineering Liaoning Petrochemical University Fushun China

ISBN: (数字)9798350387780

ISBN: (纸本)9798350387797

With the growing urgency of environmental protection, the disassembly line balancing problem has garnered significant attention. Considering some products cannot be reused even after disassembly, this paper introduces a disassembly line balancing model for U-shaped lines that incorporates destructive disassembly modes. The disassembly process is described using an AND/OR graph. Two objectives are taken into account to maximize the total profit and minimize idle time. A constrained decomposition approach with grids for evolutionary multi-objective optimization with a destructive probability generation operator is developed to solve the problem. The practicality of this algorithm is validated through a series of experiments involving products of varying sizes. Furthermore, a comprehensive evaluation is conducted by comparing it with three peer algorithms. Numerical results convincingly demonstrate the proposed model’s effectiveness and superiority.

关键词： Mathematical models Encoding Numerical models Matrix decomposition Protection Optimization

来源：评论

学校读者我要写书评

暂无评论

引用

Canadian Conference on Electrical and computer engineering (CCECE)

作者： Mohsen Tajdinian Behzad Behdani Harold R. Chamorro Vijay K. Sood Hitachi Energy Research Center Hitachi Energy Västerås Sweden Department of Electrical Sustainable Energy Delft University of Technology Delft The Netherlands School of Electrical Engineering and Computer Science KTH Royal Institute of Technology Stockholm Sweden Department of Electrical Computer and Software Engineering Faculty of Engineering and Applied Science Ontario Tech University Ontario Canada

ISBN: (数字)9798350371628

ISBN: (纸本)9798350371635

Failures of Induction Motors (IMs) can lead to unscheduled downtime and interruption in industry processes. This paper concentrates on the detection of the stator’s inter-turn faults which are one of the most frequent causes of failures in IMs. The proposed detection method is based on a similarity index that uses the current waveform. To be more specific, the proposed algorithm presents a full-cycle sliding-window-based index based on cosine similarity that only uses current signals for detection of the stator’s inter-turn faults. The proposed index cuts the phase difference before/after the disturbance and, as a result, it only depends on the size variations of the current waveform. The proposed method is technically unaffected by non-fault transient conditions including voltage imbalance, voltage sag, voltage swell, and heavy load changes. The performance of the proposed method is validated with numerous simulated scenarios and has good accuracy and speed of convergence.

关键词： Industries Voltage fluctuations Induction motors Accuracy Fault detection Power quality Stators

来源：评论

学校读者我要写书评

暂无评论

Unlearnable 3D point clouds: class-wise transformation is all you need 24

Unlearnable 3D point clouds: class-wise transformation is al...

引用

Proceedings of the 38th International Conference on Neural Information Processing Systems

作者： Xianlong Wang Minghui Li Wei Liu Hangtao Zhang Shengshan Hu Yechao Zhang Ziqi Zhou Hai Jin National Engineering Research Center for Big Data Technology and System and Services Computing Technology and System Lab and Hubei Engineering Research Center on Big Data Security and Hubei Key Laboratory of Distributed System Security and School of Cyber Science and Engineering Huazhong University of Science and Technology School of Software Engineering Huazhong University of Science and Technology Hubei Engineering Research Center on Big Data Security and Hubei Key Laboratory of Distributed System Security and School of Cyber Science and Engineering Huazhong University of Science and Technology National Engineering Research Center for Big Data Technology and System and Services Computing Technology and System Lab and Cluster and Grid Computing Lab and School of Computer Science and Technology Huazhong University of Science and Technology

ISBN: (纸本)9798331314385

Traditional unlearnable strategies have been proposed to prevent unauthorized users from training on the 2D image data. With more 3D point cloud data containing sensitivity information, unauthorized usage of this new type data has also become a serious concern. To address this, we propose the first integral unlearnable framework for 3D point clouds including two processes: (i) we propose an unlearnable data protection scheme, involving a class-wise setting established by a category-adaptive allocation strategy and multi-transformations assigned to samples; (ii) we propose a data restoration scheme that utilizes class-wise inverse matrix transformation, thus enabling authorized-only training for unlearnable data. This restoration process is a practical issue overlooked in most existing unlearnable literature, i.e., even authorized users struggle to gain knowledge from 3D unlearnable data. Both theoretical and empirical results (including 6 datasets, 16 models, and 2 tasks) demonstrate the effectiveness of our proposed unlearnable framework. Our code is available at https://***/CGCL-codes/UnlearnablePC.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Towards Robust Scene Text Image Super-resolution via Explicit Location Enhancement

arXiv

引用

arXiv 2023年

作者： Guo, Hang Dai, Tao Meng, Guanghao Xia, Shu-Tao Tsinghua Shenzhen International Graduate School Tsinghua University China College of Computer Science and Software Engineering Shenzhen University China Peng Cheng Laboratory Shenzhen China

Scene text image super-resolution (STISR), aiming to improve image quality while boosting downstream scene text recognition accuracy, has recently achieved great success. However, most existing methods treat the foreground (character regions) and background (non-character regions) equally in the forward process, and neglect the disturbance from the complex background, thus limiting the performance. To address these issues, in this paper, we propose a novel method LEMMA that explicitly models character regions to produce high-level text-specific guidance for super-resolution. To model the location of characters effectively, we propose the location enhancement module to extract character region features based on the attention map sequence. Besides, we propose the multi-modal alignment module to perform bidirectional visual-semantic alignment to generate high-quality prior guidance, which is then incorporated into the super-resolution branch in an adaptive manner using the proposed adaptive fusion module. Experiments on TextZoom and four scene text recognition benchmarks demonstrate the superiority of our method over other state-of-the-art methods. Code is available at https://***/csguoh/LEMMA. Copyright © 2023, The Authors. All rights reserved.

关键词： Location

来源：评论

学校读者我要写书评

暂无评论

Takagi-Sugeno Modeling for missing value imputations based on RReliefF Iterative Learning 2021

Takagi-Sugeno Modeling for missing value imputations based o...

引用

5th International Conference on Compute and Data Analysis, ICCDA 2021

作者： Lu, Yidan Lai, Xiaochen Zhang, Liyong Song, Juchao Zhang, Zheng Gao, Dapeng School of Software Dalian University of Technology Dalian China School of Control Science and Engineering Dalian University of Technology Dalian China International School of Information Science and Engineering Dalian University of Technology Dalian China Dut Computer Control Engineering Co Ltd. Dalian China

ISBN: (纸本)9781450389112

The presence of missing values in incomplete dataset increases the difficulty of data mining. This paper, therefore, proposes a Takagi-Sugeno fuzzy model based on RReliefF Iterative Learning (TS-RIL) method for missing value imputations. Since the regression relationship between attributes is usually different in different subsets, the premise part of this method divides input space into several fuzzy subsets based on the Fuzzy C-Means clustering with partial distance strategy (FCM-PDS). To further improve the fineness of the model, the RReliefF algorithm for each fuzzy subset is used to select the relevant features of model input. Focusing on the incomplete model input, this method adopts the iterative learning method and regards missing values as variables, so that missing values, model structure, and parameters can be learned dynamically until the iteration converges. The imputation will be completed with the accompany of the end of the iteration process. Experiments on UCI datasets are conducted to demonstrate the effectiveness of the proposed method. © 2021 ACM.

关键词： Data mining

来源：评论

学校读者我要写书评

暂无评论

GCNSDA: Predicting snoRNA-disease associations via graph convolutional network

GCNSDA: Predicting snoRNA-disease associations via graph con...

引用

2021 IEEE International Conference on Bioinformatics and Biomedicine, BIBM 2021

作者： Liu, Dayun Luo, Yi Zheng, Jingjing Xu, Hanlin Zhang, Jiaxuan Deng, Lei Central South University School of Computer Science and Engineering Changsha China Xinjiang University School of Software Urumqi China University of California San Diego Department of Cognitive Science La Jolla United States

ISBN: (纸本)9781665401265

Small nucleolar RNAs(snoRNAs) represent an abundant group of noncoding RNAs in the nucleolus of eukaryotes. Recent studies revealed that snoRNAs play a significant role in a wide range of diseases. Identifying snoRNA-disease associations can provide great insights into understanding the disease and treatment and boosting drug discovery development. Traditional methods of using biological experiments are often usually small-scale and time-consuming. Therefore, it is urgent to develop a computational framework to predict snoRNA-disease associations. In this work, we proposed a novel Graph Convolutional Network(GCN) based framework GCNSDA for predicting snoRNA-disease associations. To our best knowledge, GCNSDA is the first framework that uses a graph convolutional network to predict snoRNA-disease associations. GCNSDA is built on the bipartite graph of snoRNAs and diseases;it uses a graph neural network to discover latent factors that cause the association between snoRNAs and diseases and then generate embedded representations of snoRNAs and diseases. Experimental results showed that GCNSDA achieved better performance than the other six state-of-the-art methods. Case study further confirmed the effectiveness of GCNSDA in predicting snoRNA-disease associations. © 2021 IEEE.

关键词： Graph neural networks

来源：评论

学校读者我要写书评

暂无评论

Feature Modeling and Dimensionality Reduction to Improve ML-Based DDOS Detection Systems in SDN Environment

Feature Modeling and Dimensionality Reduction to Improve ML-...

引用

International Conference on Wavelet Active Media Technology and Information Processing (ICWAMTIP)

作者： Mohamed Ali Setitra Ilyas Benkhaddra Zine El Abidine Bensalem Mingyu Fan School of Computer Science and Engineering (Cyberspace Security) University of Electronic Science and Technology of China Chengdu China School of Information and Software Engineering University of Electronic Science and Technology of China Chengdu China

ISBN: (纸本)9781665493901

Distributed Denial of Service (DDoS) attacks are one of the most significant challenges in network security, especially in the software-Defined Network (SDN) environment, due to the centralized network management provided by the Control Plane. Considering the insufficiency of traditional detection approaches because of the growth and sophistication of DDoS attacks, exploiting Machine Learning (ML) techniques is in high demand. For this, feature modeling is essential to obtain an effective ML-based DDoS detection system, especially in the pre-processing phase. In this paper, we proposed and implemented a pre-processing model based on deep studying the dataset, going so far as to increase the features number for a better representation and, if necessary, minimize the data dimension by exploring some dimensionality reduction techniques such as Principal Component Analysis (PCA) or t-distributed Stochastic Neighbor Embedding (t-SNE). Moreover, to invest even more in our conceptual aspect relating to SDN environments, as specified in the above-cited challenge, we have chosen to implement our proposed model using an open-source SDN dataset created specially in an SDN environment. Then, the statistical characteristics of these correlations are analyzed. In addition, eight ML techniques between supervised and unsupervised models were used in our work to detect DDoS attacks. Finally, we compared our proposed model with other existing approaches. The outcome showed that the detecting reliability is improved, and the method has a good effect on detecting DDoS attacks compared with other methods.

关键词： Dimensionality reduction Measurement Support vector machines Correlation Stochastic processes Denial-of-service attack Feature extraction

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：