检索结果-内蒙古大学图书馆

Metamorphic testing of deep neural network-based autonomous driving systems using behavioural domain adequacy

Neural Computing and Applications 2025年第9期37卷 6677-6724页

作者： Kalaee, Akram Parsa, Saeed School of Computer Engineering Iran University of Science and Technology Tehran Iran

Deep neural networks (DNNs) are crucial in autonomous driving systems (ADSs) for tasks like steering control, but model inaccuracies, biased training data, and incorrect runtime parameters can compromise their reliability. Metamorphic testing (MT) enhances reliability by generating follow-up tests from mutated DNN source inputs, identifying inconsistencies as defects. Various MT techniques for ADSs include generative/transfer models, neuron-based coverage maximization, and adaptive test selection. Despite these efforts, significant challenges remain, including the ambiguity of neuron coverage’s correlation with misbehaviour detection, a lack of focus on DNN critical pathways, inadequate use of search-based methods, and the absence of an integrated method that effectively selects sources and generates follow-ups. This paper addresses such challenges by introducing DeepDomain, a grey-box multi-objective test generation approach for DNN models. It involves adaptively selecting diverse source inputs and generating domain-oriented follow-up tests. Such follow-ups explore critical pathways, extracted by neuron contribution, with broader coverage compared to their source tests (inter-behavioural domain) and attaining high neural boundary coverage of the misbehaviour regions detected in previous follow-ups (intra-behavioural domain). An empirical evaluation of the proposed approach on three DNN models used in the Udacity self-driving car challenge, and 18 different MRs demonstrates that relying on behavioural domain adequacy is a more reliable indicator than coverage criteria for effectively guiding the testing of DNNs. Additionally, DeepDomain significantly outperforms selected baselines in misbehaviour detection by up to 94 times, fault-revealing capability by up to 79%, output diversity by 71%, corner-case detection by up to 187 times, identification of robustness subdomains of MRs by up to 33 percentage points, and naturalness by two times. The results confirm that stat

关键词： Neurons

来源：评论

学校读者我要写书评

暂无评论

A recover-then-discriminate framework for robust anomaly detection

引用

science China(Information sciences) 2025年第4期68卷 300-318页

作者： Peng XING Dong ZHANG Jinhui TANG Zechao LI School of Computer Science and Engineering Nanjing University of Science and Technology Department of Electronic and Computer Engineering The Hong Kong University of Science and Technology

Anomaly detection(AD) has been extensively studied and applied across various scenarios in recent years. However, gaps remain between the current performance and the desired recognition accuracy required for practical *** paper analyzes two fundamental failure cases in the baseline AD model and identifies key reasons that limit the recognition accuracy of existing approaches. Specifically, by Case-1, we found that the main reason detrimental to current AD methods is that the inputs to the recovery model contain a large number of detailed features to be recovered, which leads to the normal/abnormal area has not/has been recovered into its original state. By Case-2, we surprisingly found that the abnormal area that cannot be recognized in image-level representations can be easily recognized in the feature-level representation. Based on the above observations, we propose a novel recover-then-discriminate(ReDi) framework for *** takes a self-generated feature map(e.g., histogram of oriented gradients) and a selected prompted image as explicit input information to address the identified in Case-1. Additionally, a feature-level discriminative network is introduced to amplify abnormal differences between the recovered and input representations. Extensive experiments on two widely used yet challenging AD datasets demonstrate that ReDi achieves state-of-the-art recognition accuracy.

关键词： recovery network HOG prompt discriminative network self-correlation loss anomaly detection

来源：评论

学校读者我要写书评

暂无评论

Improving the Surface Roughness of Dental Implant Fixture by Considering the Size,Angle and Spraying Pressure of Sandblast Particles

引用

Journal of Bionic engineering 2024年第1期21卷 303-324页

作者： Ehsan Anbarzadeh Bijan Mohammadi School of Mechanical Engineering Iran University of Science and TechnologyTehranIran

In this study,different conditions of sandblasting on dental implant fixtures were investigated to achieve the best sandblasting conditions.18 different sandblasting conditions(Using 152 implant fixture samples)were examined,including parameters such as particle size,particle blasting pressure,and particle blasting *** surface treatment of the samples was performed using the SLA+Anodizing *** testing was performed for each of the 18 different states,and the average surface roughness of each of these was compared with each ***,a bone layer was placed on the sample with the closest average surface roughness to the standard and the least amount of aluminum oxide on its surface among the 18 different states,to confirm the accuracy and quality of the desired surface roughness by examining the bone formation process and *** results showed that state No.4(sandblast particle size:75µm,spraying pressure of sandblast particles:4 bar,sandblast particle spraying angle:30 degrees),which was prepared using the SLA+Anodizing method and had a surface roughness of 1.989µm(The percentage of Al_(2)O_(3)on the surface=6%),had the best sandblasting conditions and showed 95%cell viability and accelerated the treatment and bone formation process for about a *** simulation results,using Abaqus software,indicated that the stress distribution on the surface of the implant fixture in contact with the bone surface has increased by approximately 4.3%for state *** will help prevent loosening of the dental implant fixture over time.

关键词： Dental implants Surface treatment Sandblast Surface roughness Osseointegration

来源：评论

学校读者我要写书评

暂无评论

Augmented reality without SLAM

引用

Multimedia Tools and Applications 2024年 1-17页

作者： Gholami, Aminreza Nasihatkon, Behrooz Soryani, Mohsen School of Computer Engineering Iran University of Science and Technology Tehran Iran School of Computer Engineering K. N. Toosi University of Technology Tehran Iran

Most augmented reality (AR) pipelines typically involve the computation of the camera’s pose in each frame, followed by the 2D projection of virtual objects. The camera pose estimation is commonly implemented as SLAM (Simultaneous Localisation and Mapping) algorithm. However, SLAM systems are often limited to scenarios where the camera intrinsics remain fixed or are known in all frames. This paper presents an initial effort to circumvent the pose estimation stage altogether and directly computes 2D projections using epipolar constraints. To achieve this, we initially calculate the fundamental matrices between the keyframes and each new frame. The 2D locations of objects can then be triangulated by finding the intersection of epipolar lines in the new frame. We propose a robust algorithm that can handle situations where some of the fundamental matrices are entirely erroneous. Most notably, we introduce a depth-buffering algorithm that relies solely on the fundamental matrices, eliminating the need to compute 3D point locations in the target view. By utilizing fundamental matrices, our method remains effective even when all intrinsic camera parameters vary over time. Notably, our proposed approach achieved sufficient accuracy, even with more degrees of freedom in the solution space. © The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer Nature 2024.

关键词： Augmented reality

来源：评论

学校读者我要写书评

暂无评论

Enhanced Acceleration for Generalized Nonconvex Low-Rank Matrix Learning

引用

Chinese Journal of Electronics 2025年第1期34卷 98-113页

作者： Hengmin Zhang Jian Yang Wenli Du Bob Zhang Zhiyuan Zha Bihan Wen School of Electrical and Electronic Engineering Nanyang Technological University School of Computer Science and Engineering Nanjing University of Science and Technology School of Information Science and Engineering East China University of Science and Technology Department of Electrical and Computer Engineering University of Macau

Matrix minimization techniques that employ the nuclear norm have gained recognition for their applicability in tasks like image inpainting, clustering, classification, and reconstruction. However, they come with inherent biases and computational burdens, especially when used to relax the rank function, making them less effective and efficient in real-world scenarios. To address these challenges, our research focuses on generalized nonconvex rank regularization problems in robust matrix completion, low-rank representation, and robust matrix regression. We introduce innovative approaches for effective and efficient low-rank matrix learning, grounded in generalized nonconvex rank relaxations inspired by various substitutes for the ?0-norm relaxed functions. These relaxations allow us to more accurately capture low-rank structures. Our optimization strategy employs a nonconvex and multi-variable alternating direction method of multipliers, backed by rigorous theoretical analysis for complexity and *** algorithm iteratively updates blocks of variables, ensuring efficient convergence. Additionally, we incorporate the randomized singular value decomposition technique and/or other acceleration strategies to enhance the computational efficiency of our approach, particularly for large-scale constrained minimization problems. In conclusion, our experimental results across a variety of image vision-related application tasks unequivocally demonstrate the superiority of our proposed methodologies in terms of both efficacy and efficiency when compared to most other related learning methods.

关键词： Learning systems Image recognition Minimization Computational efficiency Complexity theory Matrix decomposition Optimization Image reconstruction Singular value decomposition Convergence

来源：评论

学校读者我要写书评

暂无评论

PriChain: Efficient Privacy-Preserving Fine-Grained Redactable Blockchains in Decentralized Settings

引用

Chinese Journal of Electronics 2025年第1期34卷 82-97页

作者： Hongchen Guo Weilin Gan Mingyang Zhao Chuan Zhang Tong Wu Liehuang Zhu Jingfeng Xue School of Computer Science and Technology Beijing Institute of Technology School of Cyberspace Science and Technology Beijing Institute of Technology School of Computer and Communication Engineering University of Science and Technology Beijing

Recently, redactable blockchain has been proposed and leveraged in a wide range of real systems for its unique properties of decentralization, traceability, and transparency while ensuring controllable on-chain data redaction. However, the development of redactable blockchain is now obstructed by three limitations, which are data privacy breaches, high communication overhead, and low searching efficiency, respectively. In this paper, we propose PriChain, the first efficient privacy-preserving fine-grained redactable blockchain in decentralized settings. PriChain provides data owners with rights to control who can read and redact on-chain data while maintaining downward compatibility, ensuring the one who can redact will be able to read. Specifically, inspired by the concept of multi-authority attribute-based encryption, we utilize the isomorphism of the access control tree, realizing fine-grained redaction mechanism, downward compatibility, and collusion resistance. With the newly designed structure, PriChain can realize O（n） communication and storage overhead compared to prior O（n2） schemes. Furthermore, we integrate multiple access trees into a tree-based dictionary, optimizing searching efficiency. Theoretical analysis proves that PriChain is secure against the chosen-plaintext attack and has competitive complexity. The experimental evaluations show that PriChain realizes 10× efficiency improvement of searching and 100× lower communication and storage overhead on average compared with existing schemes.

关键词： Resistance Access control Data privacy Dictionaries Blockchains Encryption Complexity theory Usability Protection

来源：评论

学校读者我要写书评

暂无评论

Development of a composite sandwich-structure piezoresistive pressure sensor for subtle-pressures application

引用

Defence technology(防务技术) 2025年第3期45卷 48-61页

作者： Mosayeb Shiri Nowrouz Mohammad Nouri Mohammad Riahi School of Mechanical Engineering Iran University of Science and TechnologyTehran16846-13114Iran

The Design and manufacturing of a noble piezoresistive pressure sensor(PS) for subtle pressures(<1 kPa) were presented. Meanwhile, in the studies conducted in the field of pressure sensors, the measurement of subtle pressures has received less attention. The limitations in the inherent gauge factor in silicon, have led to the development of polymer and composite resistive sensitive elements. However,in the development of resistance sensing elements, the structure of composite elements with reinforcement core has not been used. The proposed PS had a composite sandwich structure consisting of a nanocomposite graphene layer covered by layers of PDMS at the bottom and on the top coupled with a polyimide(PI) core. Various tests were performed to analyze the PS. The primary design target was improved sensitivity, with a finite-element method(FEM) utilized to simulate the stress profile over piezoresistive elements and membrane deflection at various pressures. The PS manufacturing process is based on Laser-engraved graphene(LEG) technology and PDMS casting. Experimental data indicated that the manufactured PS exhibits a sensitivity of 67.28 mV/kPa for a pressure range of 30-300 Pa in ambient temperature.

关键词： Piezoresistive pressure sensor Sensor manufacturing FEM Stretchable sensor LEG

来源：评论

学校读者我要写书评

暂无评论

Aligning enhanced feature representation for generalized zero-shot learning

引用

science China(Information sciences) 2025年第2期68卷 74-88页

作者： Zhiyu FANG Xiaobin ZHU Chun YANG Hongyang ZHOU Jingyan QIN Xu-Cheng YIN School of Computer & Communication Engineering University of Science and Technology Beijing

Constructing an effective common latent embedding by aligning the latent spaces of cross-modal variational autoencoders（VAEs） is a popular strategy for generalized zero-shot learning（GZSL）. However, due to the lack of fine-grained instance-wise annotations, existing VAE methods can easily suffer from the posterior collapse problem. In this paper, we propose an innovative asymmetric VAE network by aligning enhanced feature representation（AEFR） for GZSL. Distinguished from general VAE structures, we designed two asymmetric encoders for visual and semantic observations and one decoder for visual reconstruction. Specifically, we propose a simple yet effective gated attention mechanism（GAM） in the visual encoder for enhancing the information interaction between observations and latent variables, alleviating the possible posterior collapse problem effectively. In addition, we propose a novel distributional decoupling-based contrastive learning（D2-CL） to guide learning classification-relevant information while aligning the representations at the taxonomy level in the latent representation space. Extensive experiments on publicly available datasets demonstrate the state-of-the-art performance of our method. The source code is available at https://***/seeyourmind/AEFR.

关键词： generalized zero-shot learning gated attention mechanism contrastive learning multi-modal alignment

来源：评论

学校读者我要写书评

暂无评论

Visual creativity through concept combination using quantum cognitive models

引用

Scientia iranica 2024年第2期31卷 118-135页

作者： Tabriz, M. Ahrabi Atani, T. Rafiei Ashtiani, M. Jahed-Motlagh, M.R. Laboratory School of Computer Engineering Iran University of Science and Technology Tehran16846-13114 Iran

Computational creativity modeling, including concept combination, enables us to foster deeper abilities of Artificial Intelligence (Al) agents. All hough concept combination lias been addressed in a lot of computational creativity studies, findings show incompatibility amongst empirical data of concept combination and the results of the used methods. In addition, even though recent neuroscientific studies show the crucial impact of retrieving concepts’ relations explicitly stored in episodic memory, it has been underestimated in modeling creative processes. In this paper, a quant urn cognition-based approach is used to consider the context more effectively and resolve logical inconsistencies. Also, episodic memory is leveraged as the basis for the concept combination modeling process based on the created context. The result of the proposed process is a set of meaningful concepts and expressions as a combination of stimuli and related episodes, which are used to depict a visual collage as an image. The significant improvement in the quality of results in comparison with the existing methods suggests that quantum-like modeling can be considered the foundation for developing Al agents capable of creating artistic images or assisting a person during a creative process. © 2024 Sharif university of technology.

关键词： Artificial intelligence

来源：评论

学校读者我要写书评

暂无评论

FEM and ANN analyses of short and tall geosynthetic reinforced soil walls:Evaluation of various effective aspects and ASSHTO acceptable design methods

引用

Frontiers of Structural and Civil engineering 2024年第10期18卷 1576-1594页

作者： Hamid Reza RAZEGHI Ali Akbar Heshmati RAFSANJANI Seyed Meisam ALAVI School of Civil Engineering Iran University of Science and TechnologyTehran 13114-16846Iran

In addition to confined investigations on tall geosynthetic reinforced soil(GRS)walls,a remarkable database of such walls must be analyzed to diminish engineers’concerns regarding the American Association of State Highway and Transportation Officials(AASHTO)Simplified or Simplified Stiffness Method in *** are also uncertainties regarding reinforcement load distributions of GRS walls at the ***,the current study has implemented a combination of finite element method(FEM)and artificial neural network(ANN)to distinguish the performance of short and tall GRS walls and assess the AASHTO design methods based on 88 FEM and 10000 ANN *** were conspicuous differences between the effectiveness of stiffness(63%),vertical spacing(22%),and length of reinforcements(14%)in the behavior of short and tall walls,along with predictions of geogrid load *** differences illustrated that using the Simplified Method may exert profound repercussions because it does not consider wall ***,the Simplified Stiffness Method(which incorporates wall height)predicted the reinforcement load distributions at backfill and connections ***,a Multilayer Perceptron(MLP)algorithm with a low average overall relative error(up to 2.8%)was developed to propose upper and lower limits of reinforcement load distributions,either at backfill or connections,based on 990000 ANN predictions.

关键词： geosynthetics short and tall GRS walls artificial intelligence MLP AASHTO Simplified and Simplified Stiffness Methods

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：