检索结果-内蒙古大学图书馆

National Science Review 2023年第6期10卷 13-15页

作者： Yinpeng Dong Chang Liu Wenzhao Xiang Hang Su Jun Zhu Department of Computer Science and Technology Institute for AI Tsinghua-Bosch Joint ML Center THBI Lab BNRist Center Tsinghua University Institute of Image Communication and Networks Engineering in the Department of Electronic Engineering (EE) Shanghai Jiao Tong University Key Laboratory of Intelligent Information Processing of Chinese Academy of Sciences (CAS) Institute of Computing Technology CAS Peng Cheng Laboratory Pazhou Laboratory (Huangpu)

PROBLEM In recent years,the rapid development of artificial intelligence (AI) technology,especially machine learning and deep learning, is profoundly changing human production and *** various fields,such as robotics,face recognition,autonomous driving and healthcare,AI is playing an important ***,although AI is promoting the technological revolution and industrial progress,its security risks are often *** studies have found that the wellperforming deep learning models are extremely vulnerable to adversarial examples [1-3].The adversarial examples are crafted by applying small,humanimperceptible perturbations to natural examples,but can mislead deep learning models to make wrong *** vulnerability of deep learning models to adversarial examples can raise security and safety threats to various realworld applications.

关键词： robustness adversarial example adversarial training deep learning

来源：评论

学校读者我要写书评

暂无评论

Kernel-Whitening: Overcome Dataset Bias with Isotropic Sentence Embedding

Kernel-Whitening: Overcome Dataset Bias with Isotropic Sente...

引用

2022 Conference on Empirical Methods in Natural Language processing, EMNLP 2022

作者： Gao, Songyang Dou, Shihan Zhang, Qi Huang, Xuanjing School of Computer Science Fudan University Shanghai China Shanghai Key Laboratory of Intelligent Information Processing Shanghai China

Dataset bias has attracted increasing attention recently for its detrimental effect on the generalization ability of fine-tuned models. The current mainstream solution is designing an additional shallow model to pre-identify biased instances. However, such two-stage methods scale up the computational complexity of training process and obstruct valid feature information while mitigating bias. To address this issue, we utilize the representation normalization method which aims at disentangling the correlations between features of encoded sentences. We find it also promising in eliminating the bias problem by providing isotropic data distribution. We further propose Kernel-Whitening, a Nyström kernel approximation method to achieve more thorough debiasing on nonlinear spurious correlations. Our framework is end-to-end with similar time consumption to fine-tuning. Experiments show that KernelWhitening significantly improves the performance of BERT on out-of-distribution datasets while maintaining in-distribution accuracy. © 2022 Association for Computational Linguistics.

关键词： Computational linguistics

来源：评论

学校读者我要写书评

暂无评论

DualGGAN: A New Facial Style Transfer Network 5

DualGGAN: A New Facial Style Transfer Network

引用

2023 5th International Conference on Artificial Intelligence Technologies and Applications, ICAITA 2023

作者： Zhao, Yu Peng, Chunyan Zhang, Xiaojuan School of Computer Science Qinghai Normal University Xi'ning810016 China State Key Laboratory of Intelligent Processing of Tibetan Language Qinghai Xi'ning810016 China

In response to the background penetration problem of unsupervised style transfer algorithms in most cases, a Transformer style transfer network DualGGAN based on dual generators and fusion of relative position encoding is proposed. The network is trained using the least squares generative adversarial network, and the neural network is used as the image feature extractor to generate feature maps to obtain facial image features with attention weights from the feature maps, utilizing relative position encoding and mask loss to jointly constrain feature region style transfer. The experimental results show that the DualGGAN network effectively reduces artifact generation when implementing facial style transfer, maintains good background consistency, and has good generalization ability. Experiments have shown that the FID and KID indicators on the cat2dog and older2adult datasets are significantly improved compared to other algorithms. © Published under licence by IOP Publishing Ltd.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

FIMD: Fusion-Inspired Modality Distillation for Enhanced MRI Segmentation in Incomplete Multi-Modal Scenarios 24

FIMD: Fusion-Inspired Modality Distillation for Enhanced MRI...

引用

7th International Conference on Image and Graphics processing, ICIGP 2024

作者： Yang, Rui Wang, Xiao Xu, Xin Hubei Province Key Laboratory of Intelligent Information Processing and Real-Time Industrial System Wuhan University of Science and Technology China School of Computer Science and Technology Wuhan University of Science and Technology China Hubei Provincial Key Laboratory of Multimedia and Network Communication Engineering Wuhan China

ISBN: (纸本)9798400716720

Magnetic Resonance Imaging (MRI) is an invaluable tool for brain tumor segmentation. However, in clinical practice, certain modalities might be unavailable, leading to potential performance degradation in prediction tasks. According to current implementations, different modalities are treated as independent entities during the feature extraction phase, ignoring their inherent complementary nature. In this paper, inspired by knowledge distillation techniques, we introduce the Fusion-Inspired Modality Distillation (FIMD) framework. FIMD leverages the output from a fused multi-modal model as a teacher to guide the training of individual modalities. Each modality benefits from the collective knowledge of all modalities, enhancing its performance. Our FIMD method, devoid of specific architectural constraints, seamlessly integrates into existing multi-modal brain tumor segmentation frameworks. Notably, extensive experiments on BraTS2020, BraTS2018, and BraTS2015 datasets indicate that FIMD can enhance the performance of current state-of-the-art (SOTA) algorithms on Dice by an average of 0.55-1.60%. Especially in the case of missing modalities, the effectiveness of the FIMD framework in enhancing the robustness and accuracy of brain tumor segmentation is demonstrated. © 2024 ACM.

关键词： Semantic Segmentation

来源：评论

学校读者我要写书评

暂无评论

Privacy protection oriented on data stream computing

Privacy protection oriented on data stream computing

引用

2022 International Conference on computer Application and Information Security, ICCAIS 2022

作者： He, Renjie Zheng, Xianghan Wang, Gong School of Computer and Big Data Fuzhou University Fujian Fuzhou China The Fujian Key Laboratory of Network Computing and Intelligent Information Processing Fuzhou University Fujian Fuzhou China

ISBN: (纸本)9781510663459

Data security has always been one of the public concerns, and homomorphic encryption technology, as one of the effective means to ensure data security, has always been a hot topic of academic research. Traditional machine learning is trained in a batch environment. In practice, data owners tend to outsource data and model training tasks to third-party clouds with huge computing resources, but data outsourcing carries the risk of privacy leakage. Therefore, in order to prevent this situation, an effective method is to encrypt the data before the outsourcing, and the third-party cloud trains the machine model of the ciphertext data. This method has special requirements for encryption algorithms which should support computation directly on the ciphertext data. This paper presents a novel data flow computing privacy protection framework based on the homomorphic encryption algorithm CKKS and the flow data processing engine Flink, called SDPPF (Streaming Data Privacy Protection on Flink). The proposed framework supports the functions of data stream encryption, decryption and ciphertext computing, and expands the functions of vector point multiplication and array sum operation on the basic simultaneous operation supported by the original CKKS algorithm. This paper also selects a classical machine learning algorithm: KNN algorithm, combined with the homomorphic encryption algorithm CKKS to realize the privacy protection machine learning algorithm. © 2023 SPIE.

关键词： Cryptography

来源：评论

学校读者我要写书评

暂无评论

A Local Regression-Based SLA-Aware Energy-Efficient Virtual Machine Consolidation Algorithm in Data Centers 9

A Local Regression-Based SLA-Aware Energy-Efficient Virtual ...

引用

9th International Conference on Big Data Computing and Communications, BigCom 2023

作者： Wu, Xiaodong Wang, Ronghai Lin, Guoxin Quanzhou Normal University School of Mathmatics and Computer Science China Fujian Province University Fujian Provincial Key Laboratory of Data-Intensive Computing Key Laboratory of Intelligent Computing and Information Processing Fujian China Provincial Big Data Research Institute of Intelligent Manufacturing Quanzhou China

ISBN: (纸本)9798350331240

Virtual Machine Consolidation (VMC) is an effective means to improve the resource utilization and reduce energy consumption of data centers. However, the VMC optimization that only considers energy consumption may lead to SLA (Service-Level Agreements) violations and thus fail to guarantee the QoS (Quality of Service) provided. Therefore, it is necessary to consider both energy saving and SLA violations in the process of VMC. In this paper, we propose an energy and SLA-aware VMC algorithm that considers both energy saving and SLA violation optimization in data centers. The proposed algorithm uses the local regression method to predict the utilization of virtual machines, based on which it determines whether the host will be overloaded, so as to make the decision of virtual machine migrations. The proposed VMC algorithm is evaluated with CloudSim toolkit. Simulation results demonstrate that our algorithm has good energy saving performance and can effectively reduce SLA violation in cloud data center. Considering energy consumption and SLA violation, the ESV can be reduced by 7.57% on average and 16.36% at most compared with other algorithms. © 2023 IEEE.

关键词： Virtual machine

来源：评论

学校读者我要写书评

暂无评论

Visual Object Tracking via Cascaded RPN Fusion and Coordinate Attention

引用

computer Modeling in Engineering & Sciences 2022年第9期132卷 909-927页

作者： Jianming Zhang Kai Wang Yaoqi He Lidan Kuang School of Computer and Communication Engineering Changsha University of Science and TechnologyChangsha410114China Hunan Provincial Key Laboratory of Intelligent Processing of Big Data on Transportation Changsha University of Science and TechnologyChangsha410114China

Recently,Siamese-based trackers have achieved excellent performance in object ***,the high speed and deformation of objects in the movement process make tracking ***,we have incorporated cascaded region-proposal-network(RPN)fusion and coordinate attention into Siamese *** proposed network framework consists of three parts:a feature-extraction sub-network,coordinate attention block,and cascaded RPN *** exploit the coordinate attention block,which can embed location information into channel attention,to establish long-term spatial location dependence while maintaining channel ***,the features of different layers are enhanced by the coordinate attention *** then send these features separately into the cascaded RPN for classification and *** to the two classification and regression results,the final position of the target is *** verify the effectiveness of the proposed method,we conducted comprehensive experiments on the OTB100,VOT2016,UAV123,and GOT-10k *** with other state-of-the-art trackers,the proposed tracker achieved good performance and can run at real-time speed.

关键词： Object tracking deep learning coordinate attention cascaded RPN

来源：评论

学校读者我要写书评

暂无评论

A review of computational offloading research based on game theory in a vehicular edge computing environment 7

A review of computational offloading research based on game ...

引用

7th International Conference on Mechatronics and intelligent Robotics, ICMIR 2023

作者： Qu, Zijian Tong, Yinghua Li, Maocuo School of Computer Qinghai Normal University Xining810008 China The State Key Laboratory of Tibetan Intelligent Information Processing and Application Qinghai Normal University Xining810008 China

ISBN: (数字)9781510668089

ISBN: (纸本)9781510668072

A new computing paradigm that extends the capabilities of the cloud to the edge of the network, vehicular edge computing (VEC), provides vehicles with a wealth of resources that can accommodate the rising demand for in-vehicle application computing and storage resources. This is accomplished by seamlessly fusing mobile edge computing with invehicle networks (MEC). It provides low-latency, high-bandwidth, and extremely dependable computing services to vehicles and other users with limited computing resources. This paper reviews existing game theory-based task offloading mechanisms in the vehicle edge environment. The first section introduces the current research status of vehicle edge computing in terms of its features, architecture, and applications. It covers the background and significance of task offloading in vehicle edge computing, the VEC architecture, the advantages of VEC, and several alluring application scenarios. A thorough analysis of the body of prior research is then presented, followed by a categorical discussion of game theory-based computational offloading mechanisms in the context of vehicle edges. Finally, predictions for the future of technology are made. © 2023 SPIE. All rights reserved.

关键词： computer games

来源：评论

学校读者我要写书评

暂无评论

A Compact Double-Layer Wide Beam Circularly Polarized Beidou Antenna

A Compact Double-Layer Wide Beam Circularly Polarized Beidou...

引用

IEEE International Conference on Electronic Information and Communication Technology (ICEICT)

作者： Lulu Meng Sipei Min Zhixiang Huang Yingsong Li Key Laboratory of Intelligent Computing and Signal Processing Ministry of Education Anhui University China

This paper proposes and designs a microstrip antenna that can work in the B1 band for Beidou with a double-layer substrate, and realizes the circularly polarized radiation characteristics. The antenna consists of a ground plane, two substrates, and two metal patches. The first layer of the patch is a rectangular structure as well as the second layer of the patch is a T-shaped structure with two vertical branches which are to achieve circular polarization. This structure adopts a simple coaxial feeding method, with the probe passing through the second layer of the substrate to connect the ground plane and the T-shaped strip. The simulation optimization results using the HFSS 2021 show that the axial ratio (AR) beam width in both planes exceeds 160°, and the AR at the design frequency is 0.18 dB. The impedance bandwidth is 70 MHz, which can meet the requirements of Beidou's operation.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Structure-constrain Source Free Domain Adaptation with Regularized Knowledge Distillation of Source Model 13

Structure-constrain Source Free Domain Adaptation with Regul...

引用

13th IEEE Data Driven Control and Learning Systems Conference, DDCLS 2024

作者： Yang, Dan Peng, Xin Su, Cheng Tan, Dayu Chen, Chaoyang Zhong, Weimin Ministry of Education East China University of Science and Technology Key Laboratory of Smart Manufacturing in Energy Chemical Process Shanghai200237 China Ministry of Education Anhui University Key Laboratory of Intelligent Computing and Signal Processing Anhui Hefei230601 China School of Information and Electrical Engineering Hunan University of Science and Technology Xiangtan China

ISBN: (纸本)9798350361674

Transfer learning aims to leverage the knowledge of the source domain to help the learning of unlabeled target domain models. However, all conventional transfer learning methods assume that samples from source and target domains are always freely accessible. However, this assumption is rarely plausible in real-world due to data privacy. To combat this challenge, we introduce Structure-constrain Source Free Domain Adaptation with regularized Knowledge Distillation of source model (SSFDA-KD). Our method is based on the observations that a deterministic classification result can be obtained when the target data distribution aligns with the source data distribution, and a well-trained model not only yields a deterministic classification for a sample but also has clear category boundaries, which implies samples that are close together belong to the same category. From those, we first generate the source model with labeled source data, which is regarded as a well-trained model. Then, we identify target prototypes with lower self-entropy and assign the pseudo-labels to other samples according to the neighborhood structure. Finally, we refine the target model by minimizing prediction loss, constraining internal structure via triplet loss, and reducing knowledge distillation loss. Experiments of fault diagnosis based on WWTPs are conducted to validate its effectiveness. © 2024 IEEE.

关键词： Domain Knowledge

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：