检索结果-内蒙古大学图书馆

13th EAI International Conference on Mobile Multimedia Communications, Mobimedia 2020

作者： Zhang, Hao Zhou, Joey Tianyi Wang, Tianying Tsang, Ivor W. Goh, Rick Siow Mong Institute of High Performance Computing A*STAR Singapore AAII University of Technology Sydney Australia

ISBN: (纸本)9781631902710

Ensemble learning consistently improves the performance of multi-class classification through aggregating a series of base classifiers. To this end, data-independent ensemble methods like Error Correcting Output Codes (ECOC) attract increasing attention due to its easiness of implementation and parallelization. Specifically, traditional ECOCs and its general extension N-ary ECOC decompose the original multi-class classification problem into a series of independent simpler classification subproblems. Unfortunately, integrating ECOCs, especially N-ary ECOC with deep neural networks, termed as Deep N-ary ECOC, is not straightforward and yet fully exploited in the literature, due to the high expense of training base learners. To facilitate the training of N-ary ECOC with deep learning base learners, we further propose three different variants of parameter sharing architectures for deep N-ary ECOC. To verify the generalization ability of deep N-ary ECOC, we conduct experiments by varying the backbone with different deep neural network architectures for both image and text classification tasks. Furthermore, extensive ablation studies on deep N-ary ECOC show its superior performance over other deep data-independent ensemble methods. Copyright © 2020 EAI.

关键词： Learning systems

来源：评论

学校读者我要写书评

暂无评论

WeakPolyp-SAM: Segment Anything Model-driven weakly-supervised polyp segmentation

引用

Knowledge-Based Systems 2025年

作者： Yiming Zhao Tao Zhou Yunqi Gu Yi Zhou Yizhe Zhang Ye Wu Huazhu Fu School of Computer Science and Engineering Nanjing University of Science and Technology Nanjing 210094 China School of Computer Science and Engineering Southeast University Nanjing 211189 China Institute of High Performance Computing A*STAR Singapore

Accurate polyp segmentation is essential for early detection of polyps, which is of significant clinical importance for the prevention of colorectal cancer. Although various fully-supervised deep learning techniques have been developed for polyp segmentation, pixel-wise annotation by physicians during diagnosis is both time-consuming and expensive. Meanwhile, visual foundation models such as the Segment Anything Model (SAM) have demonstrated remarkable performance in general segmentation tasks. However, directly applying SAM to medical segmentation may not yield satisfactory results due to the lack of medical knowledge. In this paper, we propose a novel SAM-driven weakly-supervised polyp segmentation framework (termed WeakPolyp-SAM), which enables a collaborative learning process between our segmentation network and SAM to boost the model performance. Specifically, we propose a Cross-aware Feature Aggregation Network (CFANet) to effectively aggregate cross-level features and leverage global cues for weakly-supervised polyp segmentation. Within CFANet, we propose a Cross-level Fusion Module (CFM) that integrates the adjacent features to enhance the representation capabilities of different resolution features. Additionally, a Local-and-global Context Decoder (LCD) is presented to capture richer features across multiple levels. Moreover, we present a box-augmentation strategy that combines the segmentation maps generated by CFANet with scribble annotations to create more precise prompts. These prompts are then fed into SAM, generating segmentation SAM-driven masks, which provide additional supervision to effectively train CFANet. Furthermore, we present an image-level filtering mechanism to filter out unreliable masks. Experimental results demonstrate that our WeakPolyp-SAM outperforms state-of-the-art weakly-supervised segmentation methods. The scribble-annotated datasets and code will be released at https://***/taozh2017/WeakPolyp-SAM .

关键词： Polyp segmentation Segment Anything Model Cross-level fusion Local-and-global context decoder

来源：评论

学校读者我要写书评

暂无评论

MOVE: Effective and Harmless Ownership Verification via Embedded External Features

引用

IEEE Transactions on Pattern Analysis and Machine Intelligence 2025年第6期47卷 4734-4751页

作者： Yiming Li Linghui Zhu Xiaojun Jia Yang Bai Yong Jiang Shu-Tao Xia Xiaochun Cao Kui Ren State Key Laboratory of Blockchain and Data Security Zhejiang University Hangzhou China Tsinghua Shenzhen International Graduate School Tsinghua University Shenzhen China Nanyang Technological University Singapore College of Computing and Data Science Nanyang Technological University Singapore ByteDance Inc San Jose CA USA Research Center of Artificial Intelligence Peng Cheng Laboratory Shenzhen China School of Cyber Science and Technology Sun Yat-sen University Shenzhen China Hangzhou High-Tech Zone Binjiang Institute of Blockchain and Data Security Hangzhou China

Currently, deep neural networks (DNNs) are widely adopted in different applications. Despite its commercial values, training a well-performing DNN is resource-consuming. Accordingly, the well-trained model is valuable intellectual property for its owner. However, recent studies revealed the threats of model stealing, where the adversaries can obtain a function-similar copy of the victim model, even when they can only query the model. In this paper, we propose an effective and harmless model ownership verification (MOVE) to defend against different types of model stealing simultaneously, without introducing new security risks. In general, we conduct the ownership verification by verifying whether a suspicious model contains the knowledge of defender-specified external features. Specifically, we embed the external features by modifying a few training samples with style transfer. We then train a meta-classifier to determine whether a model is stolen from the victim. This approach is inspired by the understanding that the stolen models should contain the knowledge of features learned by the victim model. In particular, we develop our MOVE method under both glass-boxand closed-box settings and analyze its theoretical foundation to provide comprehensive model protection. Extensive experiments on benchmark datasets verify the effectiveness of our method and its resistance to potential adaptive attacks.

关键词： Adaptation models Training Predictive models Data models Analytical models Vectors Computational modeling Closed box Watermarking Glass box

来源：评论

学校读者我要写书评

暂无评论

An Innovative Image-Based Tabular Data Extraction Parallel Algorithm 1

引用

2nd International Conference on computing Analytics and Networking, ICCAN 2019

作者： Behera, Varun Nagesh Jolly Ranjan, Ashish Reza, Motahar High Performance Computing Lab National Institute of Science and Technology Berhampur761008 India

ISBN: (数字)9789811524141

ISBN: (纸本)9789811524134

We propose a novel algorithm for extracting data from images of tabular documents having a specific structure. Our proposed method is able to maintain the original table format and structure, and offers better efficiency over existing methodologies due to its high scalability and parallel architecture. The findings of this paper will increase efficiency in the data extraction process from image-based tabular data and help ease the digitization process of tabular records. © 2020, Springer Nature Singapore Pte Ltd.

关键词： Parallel algorithms

来源：评论

学校读者我要写书评

暂无评论

In situ visualization infrastructure for large scale simulations with structured meshes 11

In situ visualization infrastructure for large scale simulat...

引用

11th International Conferences on Computer Graphics, Visualization, Computer Vision and Image Processing, CGVCVIP 2017 and International Conference on Big Data Analytics, Data Mining and Computational Intelligence, BigDaCI 2017

作者： Cao, Yi Mo, Zeyao Ai, Zhiwei Zhang, Aiqing Xiao, Li Wang, Huawei High Performance Computing Center Institute of Applied Physics and Computational Mathematics CAEP Software Center for High Performance Numerical Simulation Beijing China

ISBN: (纸本)9789898533661

To resolve complex multiscale, multiphysics problems at high resolution, modern large-scale parallel scientific and engineering simulations with structured meshes have exceeded the scale of tens of billions mesh elements. In situ visualization play an important role in avoiding the slowest component of post-processing workflow on exasale architectures: accessing the file system twice with input/output bottlenecks. Tightly coupling in situ visualization with simulations requires minimum modification to existing code. Most previous in situ work has focused on achieving in situ visualization with individual simulation code with less regard to the promising parallel programming frameworks that can guarantee quick adaptation to a batch of simulation applications. In this paper, we present an in situ visualization infrastructure based on the parallel programming framework, J Adaptive Structured Meshes applications Infrastructure (JASMIN), which supports large-scale simulations with structured meshes. We describe an approach that opens up new applications for in situ visualization, the key technologies on which it depends, and our experiences of integrating our infrastructure into two domain-specific simulations. © 2017.

关键词： Parallel programming

来源：评论

学校读者我要写书评

暂无评论

Approximation Algorithms for Steiner Connected Dominating Set

引用

Journal of Computer Science & technology 2005年第5期20卷 713-716页

作者： Ya-Feng Wu Yin-Long Xu Guo-Liang Chen National High Performance Computing Center at Hefei Department of Computer Science and Technology University of Science and Technology of China Hefei 230027 P.R. China

Steiner connected dominating set （SCDS） is a generalization of the famous connected dominating set problem, where only a specified set of required vertices has to be dominated by a connected dominating set, and known to be NP- hard. This paper firstly modifies the SCDS algorithm of Guha and Khuller and achieves a worst case approximation ratio of （2 ＋ 1/（m - 1））H（min（△, k））＋O（1）, which outperforms the previous best result （c ＋ 1）H（min（△, k））＋ O（1） in the case of m ≥ 1 ＋1/（c - 1）, where c is the best approximation ratio for Steiner tree, A is the maximum degree of the graph, k is the cardinality of the set of required vertices, m is an optional integer satisfying 0 ≤ m ≤ min（△, k） and H is the harmonic function. This paper also proposes another approximation algorithm which is based on a greedy approach. The second algorithm can establish a worst case approximation ratio of 2 ln（min（△, k））＋ O（1）, which can also be improved to 2 lnk if the optimal solution is greater than c·e^2c＋1/2（c＋1）.

关键词： approximation algorithm Steiner connected dominated set graph algorithm NP-hard

来源：评论

学校读者我要写书评

暂无评论

Automatic energy status controlling with dynamic voltage scaling in poweraware high performance computing cluster

Automatic energy status controlling with dynamic voltage sca...

引用

2011 12th International Conference on Parallel and Distributed computing, Applications and Technologies, PDCAT 2011

作者： Chen, Yuansheng Zeng, Yu Chinese Academy of Sciences Graduate University China High Performance Computer Research Center Institute of Computing Technology CAS China Beijing Computing Center Beijing 100012 China

ISBN: (纸本)9780769545646

Recently, more and more people are focused on the energy efficiency of HPC(high performance computing) centers. The power consumption of supercomputer listed in the first 10 of top 500 is usually more than 1Mkw. Power-aware techniques have been widely used in data centre or supercomputer centre for reducing the power consumption. It is useful for us to reduce the power consumption with frequency scalable CPU being adopted in the HPC centers. Because load unbalance is a common problem in high performance computing, we can gear the frequency of CPU down when it is idle so that the power consumption can be reduced with little performance lose. In this paper, we present a new method called AESC(Automatic Energy Status Controlling) which can control the energy status of CPU automatically. Firstly, we identify the application as different kinds of power consumption by the load balancing and communication time. Secondly, we will give a control policy for controlling the energy status based on former results. By using this policy, we can automatically gear the frequency of CPU and get the best result that 20% less energy while increasing execution time by only 1%. © 2011 IEEE.

关键词： Electric power utilization

来源：评论

学校读者我要写书评

暂无评论

A material removal and surface roughness evolution model for loose abrasive polishing of free form surfaces

引用

International Journal of Abrasive technology 2014年第4期6卷 269-285页

作者： Wan, Stephen Liu, Yuchan Woon, Keng Soon Tnay, Guan Leong Institute of High Performance Computing 1 Fusionolopolis Way 16-16 Connexis138632 Singapore Singapore Institute of Manufacturing Technology 71 Nanyang Drive 638075 Singapore

We report on the development of a simple process model, motivated by a need to predict the evolution of surface roughness and geometrical change over external, three dimensional free form surfaces undergoing a loose abrasive polishing process. To this end, the process model was derived from tribological wear laws and cast in terms of the point sliding velocity and apparent pressure and then benchmarked against data drawn from experiments in three disparate loose abrasive finishing techniques, namely, vibratory finishing, abrasive flow machining and magnetorheological finishing. We conclude with a brief discussion on the potential and limitations of the model in analysing loose abrasive finishing processes. Copyright © 2014 Inderscience Enterprises Ltd.

关键词： Surface roughness

来源：评论

学校读者我要写书评

暂无评论

A Novel Multiparty Quantum Secret Sharing Scheme of Secure Direct Communication Based on Bell States and Bell Measurements

引用

Chinese Physics Letters 2011年第5期28卷 22-25页

作者： SHI Run-Hua HUANG Liu-Sheng YANG Wei ZHONG Hong National High Performance Computing Center at Hefei Department of Computer Science and TechnologyUniversity of Science and Technology of ChinaHefei 230027 School of Computer Science and Technology Anhui UniversityHefei 230039 Suzhou Institute for Advanced Study University of Science and Technology of ChinaSuzhou 215123

We present a novel quantum secret sharing scheme of secure direct communication and analyze its *** scheme takes Einstein-Podolsky-Rosen(EPR)pairs in Bell states as quantum *** order to obtain thedirect communication message,all agents only need to perform Bell measurements,not to perform any local unitary *** total efficiency in this scheme approaches 100%as the classical information exchanged is unnecessary except for the eavesdropping checks.

关键词： scheme quantum unitary

来源：评论

学校读者我要写书评

暂无评论

Dependence of switching process on the perpendicular magnetic anisotropy constant in P-MTJ

引用

Chinese Physics B 2018年第9期27卷 635-638页

作者： Mao-Sen Yang Liang Fang Ya-Qing Chi State Key Laboratory of High-Performance Computing College of ComputerNational University of Defense TechnologyChangsha 410073China institute of Microelectronics College of ComputerNational University of Defense TechnologyChangsha 410073China

We investigate the dependence of the switching process on the perpendicular magnetic anisotropy （PMA） constant in perpendicular spin transfer torque magnetic tunnel junctions （P-MTJs） using micromagnetic simulations. It is found that the final stable states of the magnetization distribution of the free layer after switching can be divided into three different states based on different PMA constants： vortex, uniform, and steady. Different magnetic states can be attributed to a trade-off among demagnetization, exchange, and PMA energies. The generation of the vortex state is also related to the non-uniform stray field from the polarizer, and the final stable magnetization is sensitive to the PMA constant. The vortex and uniform states have different switching processes, and the switching time of the vortex state is longer than that of the uniform state due to hindrance by the vortex.

关键词： magnetic tunnel junction perpendicular magnetic anisotropy vortex state micromagnetic simula-tion

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：