检索结果-内蒙古大学图书馆

Compact Visual data Representation for Green Multimedia - A Human Visual System Perspective

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Chen, Peilin Fang, Xiaohan Wang, Meng Wang, Shiqi Ma, Siwei Department of Computer Science City University of Hong Kong Hong Kong The School of Data Science Lingnan University Hong Kong National Engineering Research Center of Visual Technology School of Computer Science Peking University Beijing100871 China

The Human Visual System (HVS), with its intricate sophistication, is capable of achieving ultra-compact information compression for visual signals. This remarkable ability is coupled with high generalization capability and energy efficiency. By contrast, the state-of-the-art Versatile Video Coding (VVC) standard achieves a compression ratio of around 1,000 times for raw visual data. This notable disparity motivates the research community to draw inspiration to effectively handle the immense volume of visual data in a green way. Therefore, this paper provides a survey of how visual data can be efficiently represented for green multimedia, in particular when the ultimate task is knowledge extraction instead of visual signal reconstruction. We introduce recent research efforts that promote green, sustainable, and efficient multimedia in this field. Moreover, we discuss how the deep understanding of the HVS can benefit the research community, and envision the development of future green multimedia technologies. © 2024, CC BY.

关键词： Green development

Intersecting-boundary-sensitive fingerprinting for tampering detection of DNN models 24

学校读者我要写书评

暂无评论

Intersecting-boundary-sensitive fingerprinting for tampering...

Proceedings of the 41st International Conference on Machine Learning

作者： Xiaofan Bai Chaoxiang He Xiaojing Ma Bin Benjamin Zhu Hai Jin 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 Hubei Engineering Research Center on Big Data Security and Hubei Key Laboratory of Distributed System Security Microsoft School of Computer Science and Technology Huazhong University of Science and Technology and National Engineering Research Center for Big Data Technology and System and Services Computing Technology and System Lab and Cluster and Grid Computing Lab.

Cloud-based ai services offer numerous benefits but also introduce vulnerabilities, allowing for tampering with deployed DNN models, ranging from injecting malicious behaviors to reducing computing resources. Fingerprint samples are generated to query models to detect such tampering. In this paper, we present Intersecting-Boundary-Sensitive Fingerprinting (IBSF), a novel method for black-box integrity verification of DNN models using only top-1 labels. Recognizing that tampering with a model alters its decision boundary, IBSF crafts fingerprint samples from normal samples by maximizing the partial Shannon entropy of a selected subset of categories to position the fingerprint samples near decision boundaries where the categories in the subset intersect. These fingerprint samples are almost indistinguishable from their source samples. We theoretically establish and confirm experimentally that these fingerprint samples' expected sensitivity to tampering increases with the cardinality of the subset. Extensive evaluation demonstrates that IBSF surpasses existing state-of-the-art fingerprinting methods, particularly with larger subset cardinality, establishing its state-of-the-art performance in black-box tampering detection using only top-1 labels. The IBSF code is available at: https://***/CGCL-codes/IBSF.

关键词：

Enhancing Visual Representation with Textual Semantics: Textual Semantics-Powered Prototypes for Heterogeneous Federated Learning

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Wu, Xinghao Niu, Jianwei Liu, Xuefeng Zhu, Guogang Zhang, Jiayuan Tang, Shaojie School of Computer Science and Engineering Beihang University Beijing China Zhongguancun Laboratory Beijing China Department of Management Science and Systems School of Management Center for AI Business Innovation University at Buffalo BuffaloNY United States

Federated Prototype Learning (FedPL) has emerged as an effective strategy for handling data heterogeneity in Federated Learning (FL). In FedPL, clients collaboratively construct a set of global feature centers (prototypes), and let local features align with these prototypes to mitigate the effects of data heterogeneity. The performance of FedPL highly depends on the quality of prototypes. Existing methods assume that larger inter-class distances among prototypes yield better performance, and thus design different methods to increase these distances. However, we observe that while these methods increase prototype distances to enhance class discrimination, they inevitably disrupt essential semantic relationships among classes, which are crucial for model generalization. This raises an important question: how to construct prototypes that inherently preserve semantic relationships among classes? Directly learning these relationships from limited and heterogeneous client data can be problematic in FL. Recently, the success of pretrained language models (PLMs) demonstrates their ability to capture semantic relationships from vast textual corpora. Motivated by this, we propose FedTSP, a novel method that leverages PLMs to construct semantically enriched prototypes from the textual modality, enabling more effective collaboration in heterogeneous data settings. We first use a large language model (LLM) to generate fine-grained textual descriptions for each class, which are then processed by a PLM on the server to form textual prototypes. To address the modality gap between client image models and the PLM, we introduce trainable prompts, allowing prototypes to adapt better to client tasks. Extensive experiments demonstrate that FedTSP mitigates data heterogeneity while significantly accelerating convergence. Copyright © 2025, The Authors. All rights reserved.

关键词： Federated learning

DG-CNN:Introducing Margin Information into Convolutional Neural Networks for Breast Cancer Diagnosis in Ultrasound Images

学校读者我要写书评

暂无评论

Journal of computer science & Technology 2022年第2期37卷 277-294页

作者： Xiao-Zheng Xie Jian-Wei Niu Senior Member Xue-Feng Liu Qing-Feng Li Yong Wang Jie Han Shaojie Tang State Key Laboratory of Virtual Reality Technology and Systems School of Computer Science and Engineering Beihang UniversityBeijing 100191China Beijing Advanced Innovation Center for Big Data and Brain Computing Beihang UniversityHangzhou 310051China Department of Diagnostic Ultrasound National Cancer CenterChinese Academy of Medical SciencesPeking Union Medical CollegeBeijing 100021China Naveen Jindal School of Management The University of Texas at DallasRichardsonTX 75080-3021U.S.A.

Although using convolutional neural networks(CNNs)for computer-aided diagnosis(CAD)has made tremendous progress in the last few years,the small medical datasets remain to be the major bottleneck in this *** address this problem,researchers start looking for information out of the medical *** efforts mainly leverage information from natural images via transfer *** recent research work focuses on integrating knowledge from medical practitioners,either letting networks resemble how practitioners are trained,how they view images,or using extra *** this paper,we propose a scheme named Domain Guided-CNN(DG-CNN)to incorporate the margin information,a feature described in the consensus for radiologists to diagnose cancer in breast ultrasound(BUS)*** DG-CNN,attention maps that highlight margin areas of tumors are first generated,and then incorporated via different approaches into the *** have tested the performance of DG-CNN on our own dataset(including 1485 ultrasound images)and on a public *** results show that DG-CNN can be applied to different network structures like VGG and ResNet to improve their *** example,experimental results on our dataset show that with a certain integrating mode,the improvement of using DG-CNN over a baseline network structure ResNet 18 is 2.17%in accuracy,1.69%in sensitivity,2.64%in specificity and 2.57%in AUC(Area Under Curve).To the best of our knowledge,this is the first time that the margin information is utilized to improve the performance of deep neural networks in diagnosing breast cancer in BUS images.

关键词： medical consensus domain knowledge breast cancer diagnosis margin map deep neural network

Generative Multi-Form Bayesian Optimization

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Guo, Zhendong Liu, Haitao Ong, Yew-Soon Qu, Xinghua Zhang, Yuzhe Zheng, Jianmin The Data Science and Artificial Intelligence Research Center School of Computer Science and Engineering Nanyang Technological University Singapore639798 Singapore The School of Energy and Power Engineering Dalian University of Technology Dalian116024 China Singapore The Computational Intelligence Lab. School of Computer Science and Engineering Nanyang Technological University Singapore639798 Singapore The School of Computer Science and Engineering Nanyang Technological University Singapore639798 Singapore

Many real-world problems, such as airfoil design, involve optimizing a black-box expensive objective function over complex structured input space (e.g., discrete space or non-Euclidean space). By mapping the complex structured input space into a latent space of dozens of variables, a two-stage procedure labeled as generative model based optimization (GMO) in this paper, shows promise in solving such problems. However, the latent dimension of GMO is hard to determine, which may trigger the conflicting issue between desirable solution accuracy and convergence rate. To address the above issue, we propose a multi-form GMO approach, namely generative multi-form optimization (GMFoO), which conducts optimization over multiple latent spaces simultaneously to complement each other. More specifically, we devise a generative model which promotes positive correlation between latent spaces to facilitate effective knowledge transfer in GMFoO. And further, by using Bayesian optimization (BO) as the optimizer, we propose two strategies to exchange information between these latent spaces continuously. Experimental results are presented on airfoil and corbel design problems and an area maximization problem as well to demonstrate that our proposed GMFoO converges to better designs on a limited computational budget. © 2025, CC BY-NC-ND.

关键词： Budget control

Spikformer V2: Join the High Accuracy Club on ImageNet with an SNN Ticket

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Zhou, Zhaokun Che, Kaiwei Fang, Wei Tian, Keyu Zhu, Yuesheng Yan, Shuicheng Tian, Yonghong Yuan, Li Peking University School of Electronic and Computer Engineering Shenzhen Graduate School China PengCheng Laboratory China Peking University School of Computer Science China Peking University Center for Data Science China Beijing Academy of Artificial Intelligence China

Spiking Neural Networks (SNNs), known for their biologically plausible architecture, face the challenge of limited performance. The self-attention mechanism, which is the cornerstone of the high-performance Transformer and also a biologically inspired structure, is absent in existing SNNs. To this end, we explore the potential of leveraging both self-attention capability and biological properties of SNNs, and propose a novel Spiking Self-Attention (SSA) and Spiking Transformer (Spikformer). The SSA mechanism eliminates the need for softmax and captures the sparse visual feature employing spike-based Query, Key, and Value. This sparse computation without multiplication makes SSA efficient and energy-saving. Further, we develop a Spiking Convolutional Stem (SCS) with supplementary convolutional layers to enhance the architecture of Spikformer. The Spikformer enhanced with the SCS is referred to as Spikformer V2. To train larger and deeper Spikformer V2, we introduce a pioneering exploration of Self-Supervised Learning (SSL) within the SNN. Specifically, we pre-train Spikformer V2 with masking and reconstruction style inspired by the mainstream self-supervised Transformer, and then finetune the Spikformer V2 on the image classification on ImageNet. Extensive experiments show that Spikformer V2 outperforms other previous surrogate training and ANN2SNN methods. An 8-layer Spikformer V2 achieves an accuracy of 80.38% using 4 time steps, and after SSL, a 172M 16-layer Spikformer V2 reaches an accuracy of 81.10% with just 1 time step. To the best of our knowledge, this is the first time that the SNN achieves 80+% accuracy on ImageNet. The code will be available at Spikformer V2. Copyright © 2024, The Authors. All rights reserved.

关键词： Image classification

Pre-Training and Fine-Tuning for Efficient Routing in Opportunistic Networks

学校读者我要写书评

暂无评论

Pre-Training and Fine-Tuning for Efficient Routing in Opport...

IEEE International Conference on Systems, Man and Cybernetics

作者： Jia Hao Xiaorui Wu Winston K. G. Seah Feng Zhang Gang Xu Ministry of Education College of Computer Science Engineering Research Center of Ecological Big Data Inner Mongolia University Hohhot China School of Engineering & Computer Science Victoria University of Wellington Wellington New Zealand School of Computer and Information Technology Shanxi University Taiyuan China

ISBN: (数字)9781665410205

ISBN: (纸本)9781665410212

In opportunistic networks, it is a challenge to find the best relay node instead of blindly selecting from among available nodes to forward messages and effectively transmit them to their destinations. By comparing the similarity between nodes, the next hop node that is most similar to the destination node is found in time. Existing node similarity-based opportunistic networks routing algorithms only calculate similarity between nodes according to nodes' properties that can be directly obtained from the network, such as the historical meeting record between nodes, etc. However, this superficial similarity calculation is obviously inadequate to describe the inherent dynamic nature of opportunistic networks at both spatial and temporal levels, and ignores the salient movement characteristics of nodes, resulting in poor routing performance. Therefore, this paper proposes an opportunistic network routing strategy based on pre-training and fine-tuning (PTFT) model. Firstly, an autoencoder is added to the graph neural network model to encode node movement behavior. Then, a pre-training graph neural network model in large scale opportunistic network scenarios is applied to learn potential features of nodes through fine-tuning. Finally, we calculate the similarity between nodes based on the node potential feature, thereby assisting the nodes to achieve an efficient routing decision. The simulation results show that our PTFT-based algorithm not only has superior performance compared to traditional routing algorithms, but also faster learning speed than other machine learning-based routing algorithms.

关键词： Representation learning Machine learning algorithms Simulation Heuristic algorithms Autoencoders Routing Feature extraction Graph neural networks Relays Cybernetics

Owl: Differential-Based Side-Channel Leakage Detection for CUDA Applications

学校读者我要写书评

暂无评论

Owl: Differential-Based Side-Channel Leakage Detection for C...

International Conference on Dependable Systems and Networks (DSN)

作者： Yu Zhao Wenjie Xue Weijie Chen Weizhong Qiang Deqing Zou Hai Jin National Engineering Research Center for Big Data Technology and System Hubei Key Laboratory of Distributed System Security Hubei Engineering Research Center on Big Data Security School of Cyber Science and Engineering Huazhong University of Science and Technology Wuhan China Jinyinhu Laboratory Wuhan China National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Lab Cluster and Grid Computing Lab School of Computer Science and Technology Huazhong University of Science and Technology Wuhan China

ISBN: (数字)9798350341058

ISBN: (纸本)9798350341065

Over the past decade, various methods for detecting side-channel leakage have been proposed and proven to be effective against CPU side-channel attacks. These methods are valuable in assisting developers to identify and patch side-channel vulnerabilities. Nevertheless, recent research has revealed the feasibility of exploiting side-channel vulnerabilities to steal sensitive information from GPU applications, which are beyond the reach of previous side-channel detection methods. Therefore, in this paper, we conduct an in-depth examination of various GPU features and present Owl, a novel side-channel detection tool targeting CUDA applications on NVIDIA GPUs. Owl is designed to detect and locate side-channel leakage in various types of CUDA applications. When tracking the execution of CUDA applications, we design a hierarchical tracing scheme and extend the A-DCFG (Attributed Dynamic Control Flow Graph) to address the massively parallel execution in CUDA, ensuring Owl's detection scalability. After completing the initial assessment and filtering, we conduct statistical tests on the differences in program traces to determine whether they are indeed caused by input variations, subsequently facilitating the positioning of side-channel leaks. We evaluate Owl's capability to detect side-channel leaks by testing it on Libgpucrypto, PyTorch, and nvJPEG. Meanwhile, we verify that our solution effectively handles a large number of threads. Owl has successfully identified hundreds of leaks within these applications. To the best of our knowledge, we are the first to implement side-channel leakage detection for general CUDA applications.

关键词： Target tracking Statistical analysis Filtering Scalability Graphics processing units Side-channel attacks Feature extraction

Video Stabilization via Prediction with Time-Series Network and Image Inpainting with Pyramid Fusion

学校读者我要写书评

暂无评论

Chinese Journal of Electronics 2021年第6期30卷 1103-1110页

作者： CHENG Keyang LI Shichao RONG Lan WANG Wenshan SHI Wenxi ZHAN Yongzhao School of Computer Science and Communication Engineering Jiangsu University Jiangsu Province Big Data Ubiquitous Perception and Intelligent Agricultural Application Engineering Research Center Cyber Space Security Academy of Jiangsu University National Engineering Laboratory for Public Safety Risk Perception and Control by Big Data Xinjiang Lianhaichuangzhi Information Technology Co. Ltd.

Due to the poor filling effect of the video image defect commonly used in the video stabilization field, the video is seemed still unstable after the image stabilization process, which seriously affects the visual effect. To solve this problem, we improve a video stabilization method based on time-series network prediction and pyramid fusion restoration is proposed to optimize the visual effect after image *** flow of the proposed method is as follows: First,it is adaptive to determine whether the defect of the corresponding frame at the current time needs padding inpainting. Then, for the frame that needs to be inpainting,the frames generated before the current moment are sent to the model combining the convolutional neural networks and the gate recurrent unit to predict the part to be filled. Next the current defect image and the complete image to be filled are brought into the Laplacian pyramid reconstruction, and the improved weighted optimal suture is introduced for splicing during the fusion. Finally, the video frame is cut after reconstruction. The method is tested on a data set composed of videos commonly used in the field of video stabilization. The experimental results show that the average peak signal to noise ratio of the method is 2 to 5 d B higher than that of the comparison algorithm, and the average structural similarity index is improved by about 2% to 7% compared with the contrast algorithm.

关键词： image restoration image sequences video frame cutting improved weighted optimal suture gate recurrent unit video signal processing video stabilization method image fusion Laplacian pyramid reconstruction padding inpainting image resolution neural nets convolutional neural networks video image defect pyramid fusion restoration image inpainting visual effect optimization image reconstruction time-series network prediction image stabilization process