检索结果-内蒙古大学图书馆

MOVE: Effective and Harmless Ownership Verification via Embedded External Features

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

Towards Sample-specific Backdoor Attack with Clean Labels via Attribute Trigger

引用

IEEE Transactions on Dependable and Secure computing 2025年

作者： Zhu, Mingyan Li, Yiming Guo, Junfeng Wei, Tao Xia, Shu-Tao Qin, Zhan Tsinghua University Tsinghua Shenzhen International Graduate School Shenzhen518055 China Nanyang Technological University College of Computing and Data Science 639798 Singapore Zhejiang University State Key Laboratory of Blockchain and Data Security Hangzhou310007 China University of Maryland Department of Computer Science College ParkMD20742 United States Ant Group Hangzhou310023 China Peng Cheng Laboratory Research Center of Artificial Intelligence Shenzhen518000 China Institute of Blockchain and Data Security Hangzhou310053 China

Currently, sample-specific backdoor attacks (SSBAs) are the most advanced and malicious methods since they can easily circumvent most of the current backdoor defenses. In this paper, we reveal that SSBAs are not sufficiently stealthy due to their poisoned-label nature, where users can discover anomalies if they check the image-label relationship. In particular, we demonstrate that it is ineffective to directly generalize existing SSBAs to their clean-label variants by poisoning samples solely from the target class. We reveal that it is primarily due to two reasons, including (1) the 'antagonistic effects' of ground-truth features and (2) the learning difficulty of sample-specific features. Accordingly, trigger-related features of existing SSBAs cannot be effectively learned under the clean-label setting due to their mild trigger intensity required for ensuring stealthiness. We argue that the intensity constraint of existing SSBAs is mostly because their trigger patterns are 'content-irrelevant' and therefore act as 'noises' for both humans and DNNs. Motivated by this understanding, we propose to exploit content-relevant features, a.k.a. (human-relied) attributes, as the trigger patterns to design clean-label SSBAs. This new attack paradigm is dubbed backdoor attack with attribute trigger (BAAT). Extensive experiments are conducted on benchmark datasets, which verify the effectiveness of our BAAT and its resistance to existing defenses. © 2025 IEEE.

关键词： AI Security Backdoor Attack Clean-label Attack Sample-specific Attack Trustworthy ML

来源：评论

学校读者我要写书评

暂无评论

Unified Domain Adaptive Semantic Segmentation

引用

IEEE Transactions on Pattern Analysis and Machine intelligence 2025年

作者： Zhang, Zhe Wu, Gaochang Zhang, Jing Zhu, Xiatian Tao, Dacheng Chai, Tianyou Northeastern University State Key Laboratory of Synthetical Automation for Process Industries Shenyang China Wuhan University School of Computer Science China University of Surrey Surrey Institute for People-Centred Artificial Intelligence Centre for Vision Speech and Signal Processing Guildford United Kingdom Nanyang Technological University College of Computing & Data Science Singapore

Unsupervised Domain Adaptive Semantic Segmentation (UDA-SS) aims to transfer the supervision from a labeled source domain to an unlabeled and shifted target domain. The majority of existing UDA-SS works typically consider images whilst recent attempts have extended further to tackle videos by modeling the temporal dimension. Although two lines of research share the major challenges - overcoming the underlying domain distribution shift, their studies are largely independent. It causes several issues: (1) The insights gained from each line of research remain fragmented, leading to a lack of holistic understanding of the problem and potential solutions. (2) Preventing the unification of methods and best practices across two scenarios (images and videos) will lead to redundant efforts and missed opportunities for cross-pollination of ideas. (3) Without a unified approach, the knowledge and advancements made in one scenario may not be effectively transferred to the other, leading to suboptimal performance and slower progress. Under this observation, we advocate unifying the study of UDA-SS across video and image scenarios, enabling a more comprehensive understanding, synergistic advancements, and efficient knowledge sharing. To that end, we explore the unified UDA-SS from a general domain augmentation perspective, serving as a unifying framework, enabling improved generalization, and potential for cross-pollination, ultimately contributing to the practical impact and overall progress. Specifically, we propose a Quad-directional Mixup (QuadMix) method, characterized by tackling intra-domain discontinuity, fragmented gap bridging, and feature inconsistencies through four-directional paths designed for intra- and inter-domain mixing within an explicit feature space. To deal with temporal shifts within videos, we incorporate optical flow-guided feature aggregation across spatial and temporal dimensions for fine-grained domain alignment, which is extendable to image scenarios

关键词： Semantic Segmentation

来源：评论

学校读者我要写书评

暂无评论

Air Pollution Monitoring By Integrating Local and Global Information in Self-Adaptive Multiscale Transform Domain

引用

IEEE Transactions on Multimedia 2025年

作者： Gu, Ke Liu, Yuchen Liu, Hongyan Liu, Bo Qiao, Junfei Lin, Weisi Zhang, Wenjun Beijing University of Technology School of Information Science and Technology Engineering Research Center of Intelligent Perception and Autonomous Control of Ministry of Education Beijing Laboratory of Smart Environmental Protection Beijing Key Laboratory of Computational Intelligence and Intelligent System Beijing Artificial Intelligence Institute Beijing100124 China Massey University School of Mathematical and Computational Sciences New Zealand Nanyang Technological University College of Computing and Data Science Singapore Shanghai Jiao Tong University Institute of Image Communication and Information Processing Shanghai China

As the deadliest form of pollution, air pollution had a prolonged severe damage to the human health and life safety of nearly 99% of the world's population. Facing to the problem that billions of tons of pollutants discharged into the air per year by industrial activities, this paper proposed a novel image-based air pollution monitor (IAPM) by incorporating local and global information in the self-adaptive multiscale transform domain, so as to achieve the timely and effective leakage detection of typical air pollutants from a single image. To be specific, this paper first developed a screen-shaped module according to two significant findings in visual neuroscience, which include the high sensitivity of human eyes to horizontal and vertical stimuli and the center-surround inhibition, by designing and fusing the square module, horizontal strip module and vertical strip module parallelly for simulating the behaviour of human eyes to extract local features. Second, the learnable weights and proportional mapping were applied to incorporate the screen-shaped module and lightweight vision transformer as backbone, towards more richly exploiting and fusing local and global information just as the way a brain perceives external stimuli. Third, a new self-adaptive multiscale transform domain method was devised based on two motivations from the visual characteristics of multiscale perception and the brain characteristics of self-adaptive domain transform to modify the backbone by using the operations of pooling and pointwise convolution. Extensive experiments implemented on the datasets of carbon particulate matters and ethylene leakage confirmed the superior monitoring performance of the proposed IAPM model beyond the state-of-the-art (SOTA) peers by an accuracy gain of about 4%. Furthermore, the proposed IAPM model only required 0.089 GFLOPs and 0.15 million model parameters, remarkably outperforming SOTA competitors in computational efficiency and storage resources. © 1999-2

关键词： Information leakage

来源：评论

学校读者我要写书评

暂无评论

Restoring Missing Slices of Serial Section Electron Microscopy Using Diffusion Models 2nd

Restoring Missing Slices of Serial Section Electron Microsco...

引用

2nd International Conference on Applied intelligence, ICAI 2024

作者： Yang, Hongyu Zhang, Ruobing Zhang, Qinhu Ningbo Institute of Digital Twin Eastern Institute of Technology No. 568 Tongxin Road Zhuangshi Street Zhejiang Ningbo315201 China Big Data and Intelligent Computing Research Center Guangxi Academy of Science Nanning530007 China University of Science and Technology of China No. 96 Jinzhai Road Anhui Baohe District Hefei230022 China Hefei Comprehensive National Science Center Institute of Artificial Intelligence Hefei China

ISBN: (纸本)9789819619030

Serial section electron microscopy (ssEM), an advanced three-dimensional imaging technique, has played a crucial role in studying neuronal connections and microstructures of the brain. However, imperfect sample preparation and image acquisition randomly cause degradation of ssEM images, where Missing Slices (MS) is the most significant one. The previous deep learning methods such as interpolation model using spatially-adaptive convolutions have been proven to outperform conventional approaches, but those methods fail to recover some high-frequency information of missing slices. In this work, we propose a novel approach leveraging Denoising Diffusion Probabilistic Models (DDPM) for MS restoration in ssEM images. Our method improves the U-Net backbone with asymmetric and symmetric 3D convolutions, adapting to the anisotropic characteristic of ssEM images. Then, we introduce the Adaptive and Learnable Reconstruction module (ALR) and First and Last slices Attention Block (FLAB) to extract the features of slices effectively. Moreover, we employ a diffusion correction method, reducing training-testing data discrepancies. Experiment results demonstrate the effectiveness of our method in generating realistic missing slices, surpassing previous methods. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Restoration

来源：评论

学校读者我要写书评

暂无评论

High-order diversity feature learning for pedestrian attribute recognition

引用

Neural Networks 2025年 188卷 107463页

作者： Wu, Junyi Huang, Yan Gao, Min Niu, Yuzhen Chen, Yuzhong Wu, Qiang Key Laboratory of Network Computing and Intelligent Information Processing College of Computer and Data Science Fuzhou University Fujian Fuzhou China Australian Artificial Intelligence Institute University of Technology Sydney NSW Australia Fujian Key Lab for Intelligent Processing and Wireless Transmission of Media Information College of Physics and Information Engineering Fuzhou University Fujian Fuzhou China School of Electrical and Data Engineering University of Technology Sydney NSW Australia

Pedestrian attribute recognition (PAR) involves accurately identifying multiple attributes present in pedestrian images. There are two main approaches for PAR: part-based method and attention-based method. The former relies on existing segmentation or region detection methods to localize body parts and learn corresponding attribute-specific feature from the corresponding regions, where the performance heavily depends on the accuracy of body region localization. The latter adopts the embedded attention modules or transformer attention to exploit detailed feature. However, it can focus on certain body regions but often provide coarse attention, failing to capture fine-grained details, the learned feature may also be interfered with by irrelevant information. Meanwhile, these methods overlook the global contextual information. This work argues for replacing coarse attention with detailed attention and integrating it with global contextual feature from ViT to jointly represent attribute-specific regions. To tackle this issue, we propose a High-order Diversity Feature Learning (HDFL) method for PAR based on ViT. We utilize a polynomial predictor to design an Attribute-specific Detailed Feature Exploration (ADFE) module, which can construct the high-order statistics and gain more fine-grained feature. Our ADFE module is a parameter-friendly method that provides flexibility in deciding its utilization during the inference phase. A Soft-redundancy Perception Loss (SPLoss) is proposed to adaptively measure the redundancy between feature of different orders, which can promote diverse characterization of features. Experiments on several PAR datasets show that our method achieves a new state-of-the-art (SOTA) performance. On the most challenging PA100K dataset, our method outperforms previous SOTA by 1.69% and achieves the highest mA of 84.92%. © 2025 The Authors

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

Dual prototypes contrastive learning based semi-supervised segmentation method for intelligent medical applications

引用

Engineering Applications of artificial intelligence 2025年 154卷

作者： Yue, Tianai Xu, Rongtao Wu, Jingqian Yang, Wenjie Du, Shide Wang, Changwei Johns Hopkins University Baltimore21218 United States Jinan250014 China Shandong Provincial Key Laboratory of Computing Power Internet and Service Computing Shandong Fundamental Research Center for Computer Science Jinan250014 China State Key Laboratory of Multimodal Artificial Intelligence Systems Institute of Automation Chinese Academy of Sciences Beijing100190 China The University of Hong Kong 999077 Hong Kong College of Computer and Big Data Fuzhou University Fuzhou350108 China

In medical intelligence applications, the labeling of medical data is crucial and expensive, so it becomes urgent to explore labeling-efficient ways to train applications. Semi-supervised techniques for medical image segmentation have demonstrated potential, effectively training models using scarce labeled data alongside a wealth of unlabeled data. Therefore, semi-supervised medical image segmentation is a key issue in engineering applications of medical intelligence. Consistency constraints based on prototype alignment provide an intuitively sensible way to discover valuable insights from unlabeled data that can motivate segmentation performance. In this work, we propose a Dual prototypes Contrastive Network to motivate semi-supervised medical segmentation accuracy by imposing image-level global prototype and pixel-level local prototype constraints. First, we introduce a Background-Separation Global Prototype Contrastive Learning technique that utilizes the natural mutual exclusivity of foreground and background to separate the inter-class distances and encourage the segmentation network to obtain segmentation results that are more complete and do not contain background regions. Second, we design a Cross-Consistent Local Prototype Contrastive Learning techniques to extend the perturbation consistency of the two networks to the prototype's localized response to the feature map, thereby shaping a more stable intra-class prototype space and producing accurate and robust pixel-level predictions. Finally, we comprehensively evaluate our method on mainstream semi-supervised medical image segmentation benchmarks and settings, and experimental results show that our proposed method outperforms current state-of-the-art methods. Specifically, our method achieves a Dice Coefficient score of 91.8 on the Automatic Cardiac Diagnosis Challenge dataset using only 10% labeled data training, 1.1% ahead of the second best method. Code is available at https://***/yuelily2024/DPC

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

Spiking Neural Networks for Temporal Processing: Status Quo and Future Prospects

arXiv

引用

arXiv 2025年

作者： Ma, Chenxiang Chen, Xinyi Li, Yanchen Yang, Qu Wu, Yujie Li, Guoqi Pan, Gang Tang, Huajin Tan, Kay Chen Wu, Jibin Department of Data Science and Artificial Intelligence The Hong Kong Polytechnic University Hong Kong Department of Computing The Hong Kong Polytechnic University Hong Kong Department of Electrical and Computer Engineering National University of Singapore 119077 Singapore Institute of Automation Chinese Academy of Sciences Beijing100045 China State Key Laboratory of Brain-Machine Intelligence College of Computer Science and Technology MOE Frontier Science Center for Brain Science and Brain-Machine Integration Zhejiang University Hangzhou310027 China

Temporal processing is fundamental for both biological and artificial intelligence systems, as it enables the comprehension of dynamic environments and facilitates timely responses. Spiking Neural Networks (SNNs) excel in handling such data with high efficiency, owing to their rich neuronal dynamics and sparse activity patterns. Given the recent surge in the development of SNNs, there is an urgent need for a comprehensive evaluation of their temporal processing capabilities. In this paper, we first conduct an in-depth assessment of commonly used neuromorphic benchmarks, revealing critical limitations in their ability to evaluate the temporal processing capabilities of SNNs. To bridge this gap, we further introduce a benchmark suite consisting of three temporal processing tasks characterized by rich temporal dynamics across multiple timescales. Utilizing this benchmark suite, we perform a thorough evaluation of recently introduced SNN approaches to elucidate the current status of SNNs in temporal processing. Our findings indicate significant advancements in recently developed spiking neuron models and neural architectures regarding their temporal processing capabilities, while also highlighting a performance gap in handling long-range dependencies when compared to state-of-the-art non-spiking models. Finally, we discuss the key challenges and outline potential avenues for future research. Copyright © 2025, The Authors. All rights reserved.

关键词： Spatio-temporal data

来源：评论

学校读者我要写书评

暂无评论

Multi-Agent Systems for Autonomous IoT Network Management Using Distributed Reinforcement Learning

Multi-Agent Systems for Autonomous IoT Network Management Us...

引用

International Symposium on Advanced computing and Communication (ISACC)

作者： Neelamegam G Rajaram Venkatesan Ramya SR R.S. Ramya Akshya. J M. Sundarrajan Mani Deepak Choudhry Department of Computer Science and Engineering VSB College of Engineering and Technology Coimbatore Tamil Nadu India Department of Artificial Intelligence and Data Science Builders Engineering College Tirupur Tamil Nadu India Department of Computer Science and Engineering Dr NGP Institute of Technology Coimbatore Tamil Nadu India Department of Computer Science and Engineering Sri Ramakrishna Engineering College Coimbatore Tamil Nadu India Department of Computational Intelligence SRM Institute of Science and Technology Chennai Tamil Nadu India Department of Networking and Communications SRM Institute of Science and Technology Chennai Tamil Nadu India Department of Computing Technologies SRM Institute of Science and Technology Chennai Tamil Nadu India

ISBN: (数字)9798331523893

ISBN: (纸本)9798331523909

The ever-growing complexity of IoT networks ignited by their wide scale adoption in applications such as smart cities, the industrial automation, and health care, compelled to develop sophisticated yet resource efficient and scalable network management solution. Centralized systems are traditional in architecture, which leads to higher latency and bottlenecks, while decentralized systems can distribute workloads however the lack of adaptability to real time, along with the difficulty of extracting maximum resource usage, are limitations. In recent years, reinforcement learning (RL) based approaches have taken shape as a promising alternative to facilitate the automated learning and optimisation of task management. However, there are two challenges for existing RL based methods: slow convergence, inefficient priority assignment of tasks, and poor coordination amongst agents in the case of dynamic, large scale IoT environment. To solve these challenges, in this research we propose a distributed reinforcement learning based multi agent framework for managing IoT network. The framework also integrates a task prioritization mechanism to dynamically choose an optimal task allocation solution. Further validation was performed in a simulated IoT environment where real world inspired datasets were used. The proposed framework demonstrated 30% more task throughput than centralized, 25% less latency than decentralized, and 40% more energy efficiency than current RL based systems. The system was further validated using advanced visualizations such as anomaly detection maps and resource allocation efficiency graphs, demonstrating its capacity to manage dynamic load of tasks and resource allocation. The proposed framework addresses the critical limitations of current methods, through scalable, adaptive, and energy efficient framework for managing IoT network, and accordingly helps advancing the intelligent and autonomous IoT systems.

关键词： Smart cities Reinforcement learning Medical services Throughput Dynamic scheduling Energy efficiency Real-time systems Resource management Optimization Multi-agent systems

来源：评论

学校读者我要写书评

暂无评论

AoI-Sensitive data Forwarding with Distributed Beamforming in UAV-Assisted IoT

arXiv

引用

arXiv 2025年

作者： Lang, Zifan Liu, Guixia Sun, Geng Li, Jiahui Sun, Zemin Wang, Jiacheng Leung, Victor C.M. College of Computer Science and Technology Jilin University Changchun130012 China College of Computing and Data Science Nanyang Technological University Singapore639798 Singapore Artificial Intelligence Research Institute Shenzhen MSU-BIT University Shenzhen518115 China Department of Electrical and Computer Engineering The University of British Columbia VancouverV6T 1Z4 Canada

This paper proposes a UAV-assisted forwarding system based on distributed beamforming to enhance age of information (AoI) in Internet of Things (IoT). Specifically, UAVs collect and relay data between sensor nodes (SNs) and the remote base station (BS). However, flight delays increase the AoI and degrade the network performance. To mitigate this, we adopt distributed beamforming to extend the communication range, reduce the flight frequency and ensure the continuous data relay and efficient energy utilization. Then, we formulate an optimization problem to minimize AoI and UAV energy consumption, by jointly optimizing the UAV trajectories and communication schedules. The problem is non-convex and with high dynamic, and thus we propose a deep reinforcement learning (DRL)-based algorithm to solve the problem, thereby enhancing the stability and accelerate convergence speed. Simulation results show that the proposed algorithm effectively addresses the problem and outperforms other benchmark algorithms. Copyright © 2025, The Authors. All rights reserved.

关键词： Unmanned aerial vehicles (UAV)

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：