检索结果-内蒙古大学图书馆

arXiv 2023年

作者： Yin, Zhaoxia Yin, Heng Su, Hang Zhang, Xinpeng Gao, Zhenzhe Shanghai Key Laboratory of Multidimensional Information Processing East China Normal University Shanghai Shanghai200241 China Anhui Provincial Key Laboratory of Multimodal Cognitive Computation Anhui University Anhui Hefei230000 China Department of Computer Science and Technology Tsinghua University Beijing Beijing100000 China School of Computer Science and Technology Fudan University Shanghai Shanghai200241 China

With the development of artificial intelligence, fine-tuning pretrained models has become increasingly accessible for average users, enabling them to easily adapt models to their specific tasks. As a result, the practical applications of such models have been further accelerated across various fields. Typically, foundation models are hosted on cloud servers to meet the high demand for their services. However, this exposes them to security risks, as attackers can modify them after uploading to the cloud or transferring from a local system. To address this issue, we propose an iterative decision-based fragile watermarking algorithm that transforms normal training samples into fragile samples that are sensitive to model changes. We then compare the output of sensitive samples from the original model to that of the compromised model during validation to assess the model’s completeness. The proposed fragile watermarking algorithm is an optimization problem that aims to minimize the variance of the predicted probability distribution outputed by the target model when fed with the converted sample. We convert normal samples to fragile samples through multiple iterations. Our method has some advantages: (1) the iterative update of samples is done in a decision-based black-box manner, relying solely on the predicted probability distribution of the target model, which reduces the risk of exposure to adversarial attacks, (2) the small-amplitude multiple iterations approach allows the fragile samples to perform well visually, with a PSNR of 55 dB in TinyImageNet compared to the original samples, (3) even with changes in the overall parameters of the model of magnitude 1e-4, the fragile samples can detect such changes, and (4) the method is independent of the specific model structure and dataset. We demonstrate the effectiveness of our method on multiple models and datasets, and show that it outperforms the current state-of-the-art. Copyright © 2023, The Authors. All rights reser

关键词： Probability distributions

来源：评论

学校读者我要写书评

暂无评论

Bone Marrow Cell Segmentation Based on Improved U-Net 1

引用

6th International Conference on Data Mining and Big Data, DMBD 2021

作者： Jin, Lingmin Yu, Zhaochai Fan, Haoyi Teng, Shenghua Li, Zuoyong College of Electronic and Information Engineering Shandong University of Science and Technology Qingdao266590 China Fujian Provincial Key Laboratory of Information Processing and Intelligent Control College of Computer and Control Engineering Minjiang University Fuzhou350121 China School of Information Engineering Zhengzhou University Zhengzhou450001 China

ISBN: (数字)9789811674761

ISBN: (纸本)9789811674754

Automatic segmentation of bone marrow cells plays an important role in the diagnosis of many blood diseases such as anemia and leukemia. Due to the complex morphology and wide variety of bone marrow cells, their segmentation is still a challenging task. To improve the accuracy of bone marrow cell segmentation, we propose an end-to-end U-shaped network based on the pyramid residual convolution and the attention mechanism. Specifically, the standard convolution and dilated convolution are combined as its feature encoder, which designs a pyramid residual convolution block to extract multi-scale features. Then, the attention gating mechanism is introduced into each skip connection module for fusing the shallow and deep information. Finally, the proposed method combines convolution and deconvolution for feature decoding to achieve the final segmentation of bone marrow cells. Experiments with quantitative and qualitative comparisons are carried out on a self-built bone marrow smear dataset. Experimental results demonstrate the superiority of the proposed method over several state-of-the-art methods. © 2021, Springer Nature Singapore Pte Ltd.

关键词： Image segmentation

来源：评论

学校读者我要写书评

暂无评论

Semi-supervised feature selection using sparse laplacian support vector machine 1

引用

1st International Conference on Neural Computing for Advanced Applications, NCAA 2020

作者： Zhang, Li Zheng, Xiaohan Xu, Zhiqiang School of Computer Science and Technology Joint International Research Laboratory of Machine Learning and Neuromorphic Computing Soochow University SuzhouJiangsu215006 China Provincial Key Laboratory for Computer Information Processing Technology Soochow University SuzhouJiangsu215006 China

ISBN: (数字)9789811576706

ISBN: (纸本)9789811576690

Semi-supervised feature selection is an active topic in machine learning and data mining. Laplacian support vector machine (LapSVM) has been successfully applied to semi-supervised learning. However, LapSVM cannot be directly applied to feature selection. To remedy it, we propose a sparse Laplacian support vector machine (SLapSVM) and apply it to semi-supervised feature selection. On the basis of LapSVM, SLapSVM introduces the 1-norm regularization, which means the solution of SLapSVM has sparsity. In addition, the training procedure of SLapSVM can be formulated as solving a quadratic programming problem, which indicates that the solution of SLapSVM is unique and global. SLapSVM can perform feature selection and classification at the same time. Experimental results on semi-supervised classification problems show the feasibility and effectiveness of the proposed semi-supervised learning algorithms. © Springer Nature Singapore Pte Ltd. 2020.

关键词： Support vector machines

来源：评论

学校读者我要写书评

暂无评论

Multi-Attention Gate Based U-net For Retinal Vessel Segmentation

Multi-Attention Gate Based U-net For Retinal Vessel Segmenta...

引用

IEEE International Symposium on information (IT) in Medicine and Education, ITME

作者： Yu Zhu Rui Li Shenghua Teng Xinrong Cao College of Electronic and Information Engineering Shandong University of Science and Technology Qingdao China Fujian Provincial Key Laboratory of Information Processing and Intelligent Control College of Computer and Control Engineering Minjiang University Fuzhou China

Retinal blood vessel segmentation images can be used to detect and evaluate various cardiovascular and ophthalmic diseases. However, due to the intricate vessel structures and blurred boundaries of vessels, it is a huge challenge to efficiently and accurately segment blood vessels. To deal with the above problems, this paper improves on the U-net by firstly using multi-scale feature convolution with kernels of varying size for feature extraction. Second, a non-local attention mechanism is applied to obtain richer global semantic information. Then multi-attention gate is used in the skip connection part by inputting feature maps of various scales and dimensions and selectively learning the interrelated regions, which improves the segmentation ability of the network model for the tiny structure of blood vessels. Quantitative and qualitative experimental results on two public datasets, DRIVE and CHASE_DB1, demonstrate the effectiveness of the proposed method.

关键词： Image segmentation Semantics Logic gates Blood vessels Network architecture Feature extraction Retinal vessels

来源：评论

学校读者我要写书评

暂无评论

A Network Combining CNN and Transformer for Blind Image Super-Resolution

A Network Combining CNN and Transformer for Blind Image Supe...

引用

IEEE International Symposium on information (IT) in Medicine and Education, ITME

作者： Shuhao Zhang Zuoyong Li Shenghua Teng Kun Zeng College of Electronic and Information Engineering Shandong University of Science and Technology Qingdao China Fujian Provincial Key Laboratory of Information Processing and Intelligent Control College of Computer and Control Engineering Minjiang University Fuzhou China

Blind super-resolution (SR) requires not only estimating blur kernel, but also super-resolving low-resolution image based on estimated blur kernel. Most blind SR methods use convolutional neural networks (CNNs) for kernel estimation, which cannot exploit long-range dependency within image domain, thus failing to predict blur kernel accurately. To address this issue, we propose a network combining CNN and transformer named NCCT for kernel estimation. By modeling local and non-local image priors simultaneously, NCCT outperforms other blind SR methods in terms of kernel estimation accuracy. Moreover, we design a network module named RRFDB for constructing lightweight blind SR network, which runs faster and achieves comparative SR performance with fewer parameters compared with other state-of-the-art blind SR methods.

关键词： Superresolution Education Estimation Transformers Feature extraction Convolutional neural networks Kernel

来源：评论

学校读者我要写书评

暂无评论

EWT: Efficient Wavelet-Transformer for Single Image Denoising

arXiv

引用

arXiv 2023年

作者： Li, Juncheng Cheng, Bodong Chen, Ying Gao, Guangwei Zeng, Tieyong The School of Communication & Information Engineering Shanghai University Shanghai China Jiangsu Key Laboratory of Image and Video Understanding for Social Safety Nanjing University of Science and Technology Nanjing China The School of Computer Science and Technology Xidian University Xian China The Department of Cyberspace Security Beijing Electronic Science & Technology Institute Beijing China The Institute of Advanced Technology Nanjing University of Posts and Telecommunications Nanjing China The Provincial Key Laboratory for Computer Information Processing Technology Soochow University Suzhou China The Department of Mathematics The Chinese University of Hong Kong New Territories Hong Kong

Transformer-based image denoising methods have achieved encouraging results in the past year. However, it must uses linear operations to model long-range dependencies, which greatly increases model inference time and consumes GPU storage space. Compared with convolutional neural network-based methods, current Transformer-based image denoising methods cannot achieve a balance between performance improvement and resource consumption. In this paper, we propose an Efficient Wavelet Transformer (EWT) for image denoising. Specifically, we use Discrete Wavelet Transform (DWT) and Inverse Wavelet Transform (IWT) for downsampling and upsampling, respectively. This method can fully preserve the image features while reducing the image resolution, thereby greatly reducing the device resource consumption of the Transformer model. Furthermore, we propose a novel Dual-stream Feature Extraction Block (DFEB) to extract image features at different levels, which can further reduce model inference time and GPU memory usage. Experiments show that our method speeds up the original Transformer by more than 80%, reduces GPU memory usage by more than 60%, and achieves excellent denoising results. All code will be public. © 2023, CC BY.

关键词： Image denoising

来源：评论

学校读者我要写书评

暂无评论

Deep Learning in Palmprint Recognition-A Comprehensive Survey

arXiv

引用

arXiv 2025年

作者： Gao, Chengrui Yang, Ziyuan Jia, Wei Leng, Lu Zhang, Bob Teoh, Andrew Beng Jin College of Computer Science Sichuan University Chengdu610065 China Singapore School of Computer and Information Hefei University of Technology Hefei China Jiangxi Provincial Key Laboratory of Image Processing and Pattern Recognition Nanchang Hangkong University Nanchang China Pattern Analysis and Machine Intelligence Group Department of Computer and Information Science University of Macau Taipa China School of Electrical and Electronic Engineering College of Engineering Yonsei University Seoul Korea Republic of

Palmprint recognition has emerged as a prominent biometric technology, widely applied in diverse scenarios. Traditional handcrafted methods for palmprint recognition often fall short in representation capability, as they heavily depend on researchers' prior knowledge. Deep learning (DL) has been introduced to address this limitation, leveraging its remarkable successes across various domains. While existing surveys focus narrowly on specific tasks within palmprint recognition-often grounded in traditional methodologies-there remains a significant gap in comprehensive research exploring DL-based approaches across all facets of palmprint recognition. This paper bridges that gap by thoroughly reviewing recent advancements in DL-powered palmprint recognition. The paper systematically examines progress across key tasks, including region-of-interest segmentation, feature extraction, and security/privacy-oriented challenges. Beyond highlighting these advancements, the paper identifies current challenges and uncovers promising opportunities for future research. By consolidating state-of-the-art progress, this review serves as a valuable resource for researchers, enabling them to stay abreast of cutting-edge technologies and drive innovation in palmprint recognition. Copyright © 2025, The Authors. All rights reserved.

关键词： Biometrics

来源：评论

学校读者我要写书评

暂无评论

HSR-Diff: Hyperspectral Image Super-Resolution via Conditional Diffusion Models

HSR-Diff: Hyperspectral Image Super-Resolution via Condition...

引用

International Conference on computer Vision (ICCV)

作者： Chanyue Wu Dong Wang Yunpeng Bai Hanyu Mao Ying Li Qiang Shen National Engineering Laboratory for Integrated Aero-Space-Ground-Ocean Big Data Application Technology Shaanxi Provincial Key Laboratory of Speech & Image Information Processing School of Computer Science Northwestern Polytechnical University China Department of Computer Science Faculty of Business and Physical Sciences Aberystwyth University U.K.

Despite the proven significance of hyperspectral images (HSIs) in performing various computer vision tasks, its potential is adversely affected by the low-resolution (LR) property in the spatial domain, resulting from multiple physical factors. Inspired by recent advancements in deep generative models, we propose an HSI Super-resolution (SR) approach with Conditional Diffusion Models (HSR-Diff) that merges a high-resolution (HR) multispectral image (MSI) with the corresponding LR-HSI. HSR-Diff generates an HR-HSI via repeated refinement, in which the HR-HSI is initialized with pure Gaussian noise and iteratively refined. At each iteration, the noise is removed with a Conditional Denoising Transformer (CDFormer) that is trained on denoising at different noise levels, conditioned on the hierarchical feature maps of HR-MSI and LR-HSI. In addition, a progressive learning strategy is employed to exploit the global information of full-resolution images. Systematic experiments have been conducted on four public datasets, demonstrating that HSR-Diff outperforms state-of-the-art methods.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Hybrid Data-Free Knowledge Distillation 39

Hybrid Data-Free Knowledge Distillation

引用

39th Annual AAAI Conference on Artificial Intelligence, AAAI 2025

作者： Tang, Jialiang Chen, Shuo Gong, Chen School of Computer Science and Engineering Nanjing University of Science and Technology China Key Laboratory of Intelligent Perception and Systems for High-Dimensional Information of Ministry of Education China Jiangsu Key Laboratory of Image and Video Understanding for Social Security China Center for Advanced Intelligence Project RIKEN Japan Department of Automation Institute of Image Processing and Pattern Recognition Shanghai Jiao Tong University China

ISBN: (纸本)157735897X

Data-free knowledge distillation aims to learn a compact student network from a pre-trained large teacher network without using the original training data of the teacher network. Existing collection-based and generation-based methods train student networks by collecting massive real examples and generating synthetic examples, respectively. However, they inevitably become weak in practical scenarios due to the difficulties in gathering or emulating sufficient real-world data. To solve this problem, we propose a novel method called Hybrid Data-Free Distillation (HiDFD), which leverages only a small amount of collected data as well as generates sufficient examples for training student networks. Our HiDFD comprises two primary modules, i.e., the teacher-guided generation and student distillation. The teacher-guided generation module guides a Generative Adversarial Network (GAN) by the teacher network to produce high-quality synthetic examples from very few real-world collected examples. Specifically, we design a feature integration mechanism to prevent the GAN from overfitting and facilitate the reliable representation learning from the teacher network. Meanwhile, we drive a category frequency smoothing technique via the teacher network to balance the generative training of each category. In the student distillation module, we explore a data inflation strategy to properly utilize a blend of real and synthetic data to train the student network via a classifier-sharing-based feature alignment technique. Intensive experiments across multiple benchmarks demonstrate that our HiDFD can achieve state-of-the-art performance using 120 times less collected data than existing methods. Copyright © 2025, Association for the Advancement of Artificia Intelligence (***). All rights reserved.

关键词： Personnel training

来源：评论

学校读者我要写书评

暂无评论

Operator-Adapted Evolutionary Large-Scale Multiobjective Optimization for Voltage Transformer Ratio Error Estimation 11th

Operator-Adapted Evolutionary Large-Scale Multiobjective Opt...

引用

11th International Conference on Evolutionary Multi-Criterion Optimization, EMO 2021

作者： Huang, Changwu Li, Lianghao He, Cheng Cheng, Ran Yao, Xin Guangdong Provincial Key Laboratory of Brain-Inspired Intelligent Computation Department of Computer Science and Engineering Southern University of Science and Technology Shenzhen518055 China Key Laboratory of Image Information Processing and Intelligent Control of Education Ministry of China School of Artificial Intelligence and Automation Huazhong University of Science and Technology Wuhan430074 China

ISBN: (纸本)9783030720612

Large-scale multiobjective optimization problems (LSMOPs) exist widely in real-world applications, and they are challenging for existing evolutionary algorithms due to their massive volume of search space. Despite that a number of large-scale multiobjective evolutionary algorithms (LSMOEAs) have been proposed in recent years, their effectiveness in solving LSMOPs remains unsatisfactory. One main reason is that most existing LSMOEAs may fail to balance convergence enhancement and diversity maintenance, especially for solving real-world problems. To address this issue, we propose to use a hybridized LSMOEA with adaptive operator selection (AOS) to handle real-world LSMOPs. Specifically, the proposed hybridized LSMOEA with AOS (AOS-LSMOEA) includes multiple different offspring generation and environmental selection strategies extracted from some existing LSMOEAs. Then it uses the AOS to adaptively determine the application rates of different offspring generation and environmental selection operators in an online manner. The proposed approach is capable of taking advantage of existing LSMOEAs, and the AOS enables the algorithm to choose suitable operators for solving different LSMOPs. In this study, the proposed algorithm is expected to solve the voltage transformer ratio error estimation (TREE) problems effectively. Experimental results show that AOS-LSMOEA achieves significant performance improvement due to the hybridization of different operators and the adoption of AOS method. © 2021, Springer Nature Switzerland AG.

关键词： Multiobjective optimization

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：