检索结果-内蒙古大学图书馆

arXiv 2024年

作者： Jiang, Yuncheng Feng, Chun-Mei Ren, Jinke Wei, Jun Zhang, Zixun Hu, Yiwen Liu, Yunbi Sun, Rui Tang, Xuemei Du, Juan Wan, Xiang Xu, Yong Du, Bo Gao, Xin Wang, Guangyu Zhou, Shaohua Cui, Shuguang Goh, Rick Siow Mong Liu, Yong Li, Zhen Shenzhen518172 China Shenzhen518172 China Institute of High Performance Computing Agency for Science Technology and Research Singapore138632 Singapore College of Computer Science and Software Engineering Shenzhen University Shenzhen518060 China Shenzhen Research Institute of Big Data Shenzhen518172 China South China Hospital Health Science Center Shenzhen University Shenzhen518111 China School of Computer Science Nanjing University of Posts and Telecommunications Nanjing210023 China Affiliated Hospital of North Sichuan Medical College Sichuan637000 China North Sichuan Medical College Sichuan637000 China Shenzhen518055 China School of Computer Science Wuhan University Wuhan430072 China Thuwal23955-6900 Saudi Arabia Thuwal23955-6900 Saudi Arabia Beijing University of Posts and Telecommunications Beijing100876 China School of Biomedical Engineering Suzhou Institute for Advanced Research University of Science and Technology of China Suzhou215123 China Institute of Computing Technology Chinese Academy of Sciences Beijing100190 China

Ultrasound imaging is widely used in clinical diagnosis due to its non-invasive nature and real-time capabilities. However, conventional ultrasound diagnostics face several limitations, including high dependence on physician expertise and suboptimal image quality, which complicates interpretation and increases the likelihood of diagnostic errors. Artificial intelligence (AI) has emerged as a promising solution to enhance clinical diagnosis, particularly in detecting abnormalities across various biomedical imaging modalities. Nonetheless, current AI models for ultrasound imaging face critical challenges. First, these models often require large volumes of labeled medical data, raising concerns over patient privacy breaches. Second, most existing models are task-specific, which restricts their broader clinical utility. To overcome these challenges, we present UltraFedFM, an innovative privacy-preserving ultrasound foundation model. UltraFedFM is collaboratively pre-trained using federated learning across 16 distributed medical institutions in 9 countries, leveraging a dataset of over 1 million ultrasound images covering 19 organs and 10 ultrasound modalities. This extensive and diverse data, combined with a secure training framework, enables UltraFedFM to exhibit strong generalization and diagnostic capabilities. It achieves an average area under the receiver operating characteristic curve (AUROC) of 0.927 for disease diagnosis and a dice similarity coefficient (DSC) of 0.878 for lesion segmentation. Notably, UltraFedFM surpasses the diagnostic accuracy of mid-level ultrasonographers (4–8 years of experience) and matches the performance of expert-level sonographers (10+ years of experience) in the joint diagnosis of 8 common systemic diseases. These findings indicate that UltraFedFM can significantly enhance clinical diagnostics while safeguarding patient privacy, marking a significant advancement in AI-driven ultrasound imaging for future clinical applications. Copyri

关键词： Differential privacy

来源：评论

学校读者我要写书评

暂无评论

Deep learning-based forgery attack on document images

arXiv

引用

arXiv 2021年

作者： Zhao, Lin Chen, Changsheng Huang, Jiwu the Guangdong Key Laboratory of Intelligent Information Processing Shenzhen Key Laboratory of Media Security National Engineering Laboratory for Big Data System Computing Technology College of Electronics and Information Engineering Shenzhen University Shenzhen China Shenzhen Institute of Artificial Intelligence and Robotics for Society China

With the ongoing popularization of online services, the digital document images have been used in various applications. Meanwhile, there have emerged some deep learning-based text editing algorithms which alter the textual information of an image in an end-to-end fashion. In this work, we present a low-cost document forgery algorithm by the existing deep learning-based technologies to edit practical document images. To achieve this goal, the limitations of existing text editing algorithms towards complicated characters and complex background are addressed by a set of network design strategies. First, the unnecessary confusion in the supervision data is avoided by disentangling the textual and background information in the source images. Second, to capture the structure of some complicated components, the text skeleton is provided as auxiliary information and the continuity in texture is considered explicitly in the loss function. Third, the forgery traces induced by the text editing operation are mitigated by some post-processing operations which consider the distortions from the print-and-scan channel. Quantitative comparisons of the proposed method and the exiting approach have shown the advantages of our design by reducing the about 2/3 reconstruction error measured in MSE, improving reconstruction quality measured in PSNR and in SSIM by 4 dB and 0.21, respectively. Qualitative experiments have confirmed that the reconstruction results of the proposed method are visually better than the existing approach in both complicated characters and complex texture. More importantly, we have demonstrated the performance of the proposed document forgery algorithm under a practical scenario where an attacker is able to alter the textual information in an identity document using only one sample in the target domain. The forged-and-recaptured samples created by the proposed text editing attack and recapturing operation have successfully fooled some existing document authenticat

关键词： Textures

来源：评论

学校读者我要写书评

暂无评论

TediGAN: Text-guided diverse face image generation and manipulation

arXiv

引用

arXiv 2020年

作者： Xia, Weihao Yang, Yujiu Xue, Jing-Hao Wu, Baoyuan Tsinghua Shenzhen International Graduate School Tsinghua University China Department of Statistical Science University College London United Kingdom School of Data Science Chinese University of Hongkong Shenzhen China Secure Computing Lab of Big Data Shenzhen Research Institute of Big Data Shenzhen China

In this work, we propose TediGAN, a novel framework for multi-modal image generation and manipulation with textual descriptions. The proposed method consists of three components: StyleGAN inversion module, visual-linguistic similarity learning, and instance-level optimization. The inversion module maps real images to the latent space of a well-trained StyleGAN. The visual-linguistic similarity learns the text-image matching by mapping the image and text into a common embedding space. The instancelevel optimization is for identity preservation in manipulation. Our model can produce diverse and high-quality images with an unprecedented resolution at 10242. Using a control mechanism based on style-mixing, our Tedi- GAN inherently supports image synthesis with multi-modal inputs, such as sketches or semantic labels, with or without instance guidance. To facilitate text-guided multimodal synthesis, we propose the Multi-Modal CelebA-HQ, a large-scale dataset consisting of real face images and corresponding semantic segmentation map, sketch, and textual descriptions. Extensive experiments on the introduced dataset demonstrate the superior performance of our proposed method. Code and data are available at https://***/weihaox/TediGAN. Copyright © 2020, The Authors. All rights reserved.

关键词： Semantic Segmentation

来源：评论

学校读者我要写书评

暂无评论

Renewable-to-Ammonia: Configuration Strategy and Technoeconomic Analysis

SSRN

引用

SSRN 2023年

作者： Pan, Li Li, Jiarong Huang, Jingsi An, Qi Lin, Jin Mujeeb, Asad Xu, Yanhui Li, Gengyin Zhou, Ming Wang, Jianxiao School of Electrical and Electronic Engineering North China Electric Power University Beijing102206 China Department of Electrical Engineering Tsinghua University Beijing100084 China Harvard John A. Paulson School of Engineering and Applied Sciences CambridgeMA02138 United States Department of Industrial Engineering and Management College of Engineering Peking University Beijing100871 China National Engineering Laboratory for Big Data Analysis and Applications Peking University Beijing100871 China PKU-Changsha Institute for Computing and Digital Economy Changsha410000 China

The increasing demand for nitrogen fertilizer and chemical raw materials has provided an opportunity for the ammonia (NH3) industry. Although there are many excellent reviews proposing operation control strategies and capacity configuration strategies of renewable-to-ammonia (RE2A) systems, surprisingly little attention has been devoted to topics of public interest. Therefore, this paper proposes a technoeconomic model to research the optimal capacity configuration and quantify the levelized cost of ammonia (LCOA) for RE2A. A cost model of coal-to-ammonia (C2A) is established as a benchmark to evaluate the economic feasibility of RE2A. A case study is adopted to verify the validity of the proposed approach. It is shown that the monthly NH3 output is 6-10% of the annual output, which satisfies actual industrial production. The LCOA of RE2A is 507$/t, with investments in wind turbines and electrolysers accounting for 57% and 20%, respectively, which is lower than the NH3 market price (605-650$/t) in China at present. The LCOA of RE2A will equal the LCOA of C2A with an extra carbon tax of 47.5$/t, which confirms the economic and environmental advantages of RE2A in the long run. The sensitivity analysis of RE2A in comparison with C2A provides valuable guidelines for the application of RE2A. © 2023, The Authors. All rights reserved.

关键词： Ammonia

来源：评论

学校读者我要写书评

暂无评论

A Multichannel CNN-GRU Hybrid Architecture for sEMG Gesture Recognition

A Multichannel CNN-GRU Hybrid Architecture for sEMG Gesture ...

引用

IEEE International Conference on Bioinformatics and Biomedicine (BIBM)

作者： Shouliang Song Anming Dong Jiguo Yu Yubing Han You Zhou Key Laboratory of Computing Power Network and Information Security Ministry of Education Shandong Computer Science Center (National Supercomputer Center in Jinan) Qilu University of Technology (Shandong Academy of Sciences) Jinan China School of Computer Science and Technology Qilu University of Technology (Shandong Academy of Sciences) Jinan China Big Data Institute Qilu University of Technology (Shandong Academy of Sciences) Jinan China Shandong Provincial Key Laboratory of Computer Networks Shandong Fundamental Research Center for Computer Science Jinan China Shandong Haikan New Media Research Institute Co. Ltd.

Surface electromyography (sEMG) signal is a physiological electrical signal produced by muscle contraction. Different gestures can be effectively recognized from the characteristics of the sEMG signal. Currently, convolutional neural networks (CNNs) have been widely used in sEMG gesture recognition systems due to their capabilities in acquiring spatial features of sEMG signals. However, these classical CNNs are inefficient in extracting temporal correlation that resides in the time serials of sEMG signals, which is definitely important for gesture recognition. To overcome such a drawback of traditional CNN-based gesture recognition methods, we propose a multichannel hybrid deep learning model for gesture recognition by combining the multichannel CNNs with a gated recurrent unit (GRU). Specifically, we use multiple CNNs to preprocess the original multichannel EMG signals in a one-by-one manner to obtain the spatial features in the current observing window. The outputs of the multiple CNNs are concatenated and fed to a temporal-feature extracting module, which is designed by cascading a GRU with an attention mechanism. Through the GRU, the temporal features of successive signal frames can be established, while the attention mechanism is introduced to further focus on the key information in recognizing the gestures, which is beneficial to improve the robustness and accuracy of the model. Experiments show that the recognition accuracy of the proposed method reaches 97.6% and 96.7% on the Ninapro DB2 and Ninapro DB5 datasets, respectively. Compared with the classical CNN method, the performance improvement is 2.9% and xx% higher than that of the traditona CNN model, respectively.

关键词：

来源：评论

学校读者我要写书评

暂无评论

The minority matters: a diversity-promoting collaborative metric learning algorithm 22

The minority matters: a diversity-promoting collaborative me...

引用

Proceedings of the 36th International Conference on Neural Information Processing Systems

作者： Shilong Bao Qianqian Xu Zhiyong Yang Yuan He Xiaochun Cao Qingming Huang State Key Laboratory of Information Security Institute of Information Engineering CAS and School of Cyber Security University of Chinese Academy of Sciences Key Lab. of Intelligent Information Processing Institute of Computing Technology CAS School of Computer Science and Tech. University of Chinese Academy of Sciences Alibaba Group School of Cyber Science and Technology Shenzhen Campus Sun Yat-sen University Key Lab. of Intelligent Information Processing Institute of Computing Technology CAS and School of Computer Science and Tech. University of Chinese Academy of Sciences and Key Laboratory of Big Data Mining and Knowledge Management CAS and Peng Cheng Laboratory

ISBN: (纸本)9781713871088

Collaborative Metric Learning (CML) has recently emerged as a popular method in recommendation systems (RS), closing the gap between metric learning and Collaborative Filtering. Following the convention of RS, existing methods exploit unique user representation in their model design. This paper focuses on a challenging scenario where a user has multiple categories of interests. Under this setting, we argue that the unique user representation might induce preference bias, especially when the item category distribution is imbalanced. To address this issue, we propose a novel method called Diversity-Promoting Collaborative Metric Learning (DPCML), with the hope of considering the commonly ignored minority interest of the user. The key idea behind DPCML is to include a multiple set of representations for each user in the system. Based on this embedding paradigm, user preference toward an item is aggregated from different embeddings by taking the minimum item-user distance among the user embedding set. Furthermore, we observe that the diversity of the embeddings for the same user also plays an essential role in the model. To this end, we propose a Diversity Control Regularization Scheme (DCRS) to accommodate the multi-vector representation strategy better. Theoretically, we show that DPCML could generalize well to unseen test data by tackling the challenge of the annoying operation that comes from the minimum value. Experiments over a range of benchmark datasets speak to the efficacy of DPCML.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Robust Self-Expression Learning with Adaptive Noise Perception

SSRN

引用

SSRN 2023年

作者： Wang, Yangbo Zhou, Jie Lu, Jianglin Wan, Jun Gao, Can Lin, Qingshui College of Computer Science and Software Engineering Shenzhen University Shenzhen518060 China National Engineering Laboratory for Big Data System Computing Technology Shenzhen University Shenzhen518060 China SZU Branch Shenzhen Institute of Artificial Intelligence and Robotics for Society Shenzhen518060 China SMILE Lab Department of ECE College of Engineering Northeastern University Boston02115 United States School of Information and Safety Engineering Zhongnan University of Economics and Law Wuhan430073 China Basic Teaching Department Liaoning Technical University Huludao125105 China

Self-expression learning methods often obtain a coefficient matrix to measure the similarity between pairs of samples. However, directly using the raw data to represent each sample under the self-expression framework may not be ideal, as noise points are inevitably involved in the process of representing clean samples. To address this issue, this work proposes a novel self-expression model called robust Self-Expression learning with adaptive Noise Perception (SENP). SENP decomposes each sample into a clean part and a noisy part, and samples with large self-expression losses can be recognized as the noise points. A reliable coefficient matrix can then be learned by using only the clean points to reconstruct the clean part of each sample. By simultaneously detecting the noisy part of each sample and noise points, and adaptively mitigating their negative impacts, the representative ability of the generated coefficient matrix is improved. Moreover, inspired by the solution of non-negative matrix factorization (NMF), an effective algorithm is formed to optimize SENP. Extensive experiments on well-known benchmark datasets demonstrate the superiority of SENP compared to several state-of-the-art methods. © 2023, The Authors. All rights reserved.

关键词： Non-negative matrix factorization

来源：评论

学校读者我要写书评

暂无评论

Picosecond Optospintronic Tunnel Junctions for Non-volatile Photonic Memories

Research Square

引用

Research Square 2021年

作者： Wang, Luding Cheng, Houyi Li, Pingzhi Liu, Yang van Hees, Youri L.W. Lavrijsen, Reinoud Lin, Xiaoyang Cao, Kaihua Koopmans, Bert Zhao, Weisheng Fert Beijing Institute School of Integrated Circuit Science and Engineering Beijing Advanced Innovation Center for Big Data and Brain Computing Beihang University Beijing100191 China Department of Applied Physics Institute for Photonic Integration Eindhoven University of Technology P.O. Box 513 Eindhoven5600 MB Netherlands

Perpendicular magnetic tunnel junctions are one of the building blocks for spintronic memories, which allow fast nonvolatile data access, offering substantial potentials to revolutionize the mainstream computing architecture1-3. However, conventional switching mechanisms of such devices are fundamentally hindered by spin polarized currents4, either spin transfer torque5,6 or spin orbit torque7,8 with spin precession time limitation and excessive power dissipation. These physical constraints significantly stimulate the advancement of modern spintronics1,4. Here, we report an optospintronic tunnel junction using a photonic-spintronic combination. This composite device incorporates an all-optically switchable Co/Gd bilayer coupled to a CoFeB/MgO-based perpendicular magnetic tunnel junction by the Ruderman-Kittel-Kasuya-Yosida interaction. A picosecond all-optical operation of the optospintronic tunnel junction is explicitly confirmed by time-resolved measurements. Moreover, the device shows a considerable tunnel magnetoresistance and thermal stability. This proof-of-concept device represents an essential step towards ultrafast photonic memories with THz data access, as well as ultralow power consumption. © 2021, CC BY.

关键词： Tunnel junctions

来源：评论

学校读者我要写书评

暂无评论

Salient feature extractor for adversarial defense on deep neural networks

arXiv

引用

arXiv 2021年

作者： Chen, Jinyin Chen, Ruoxi Zheng, Haibin Ming, Zhaoyan Jiang, Wenrong Cui, Chen Institute of Cyberspace Security Zhejiang University of Technology Hangzhou China College of Information Engineering Zhejiang University of Technology Hangzhou China Institute of Computing Innovation Zhejiang Univeristy Hangzhou China College of Computer Science Hangzhou Dianzi University Hangzhou China The Big Data and Cyber Security Research Institute Zhejiang Police College Hangzhou China

Recent years have witnessed unprecedented success achieved by deep learning models in the field of computer vision. However, their vulnerability towards carefully crafted adversarial examples has also attracted the increasing attention of researchers. Motivated by the observation that adversarial examples are due to the non-robust feature learned from the original dataset by models, we propose the concepts of salient feature(SF) and trivial feature(TF). The former represents the class-related feature, while the latter is usually adopted to mislead the model. We extract these two features with coupled generative adversarial network model and put forward a novel detection and defense method named salient feature extractor (SFE) to defend against adversarial attacks. Concretely, detection is realized by separating and comparing the difference between SF and TF of the input. At the same time, correct labels are obtained by re-identifying SF to reach the purpose of defense. Extensive experiments are carried out on MNIST, CIFAR-10, and ImageNet datasets where SFE shows state-of-the-art results in effectiveness and efficiency compared with baselines. Furthermore, we provide an interpretable understanding of the defense and detection process. The code of SFE could be downloaded from https://***/ haibinzheng/SFE . Copyright © 2021, The Authors. All rights reserved.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

Constraints on Primordial Magnetic Fields from the Lyman-α forest

arXiv

引用

arXiv 2025年

作者： Pavičević, Mak Iršič, Vid Viel, Matteo Bolton, James S. Haehnelt, Martin G. Martin-Alvarez, Sergio Puchwein, Ewald Ralegankar, Pranjal SISSA - International School for Advanced Studies Via Bonomea 265 TriesteI-34136 Italy IFPU Institute for Fundamental Physics of The Universe Via Beirut 2 TriesteI-34151 Italy INFN Sezione di Trieste Via Valerio 2 TriesteI-34127 Italy INAF - Osservatorio Astronomico di Trieste Via G. B. Tiepolo 11 TriesteI-34143 Italy KICC - Kavli Institute for Cosmology Cambridge Madingley Road CambridgeCB3 0HA United Kingdom ICSC - Centro Nazionale di Ricerca in High Performance Computing Big Data e Quantum Computing Via Magnanelli 2 Bologna Italy School of Physics and Astronomy University of Nottingham University Park NottinghamNG7 2RD United Kingdom Institute of Astronomy Kavli Institute for Cosmology University of Cambridge Madingley Road CambridgeCB3 0HA United Kingdom Stanford University StanfordCA94305 United States An der Sternwarte 16 Potsdam14482 Germany

We present the first constraints on primordial magnetic fields from the Lyman-α forest using full cosmological hydrodynamic simulations. At the scales and redshifts probed by the data, the flux power spectrum is extremely sensitive to the extra power induced by primordial magnetic fields in the linear matter power spectrum, at a scale that we parametrize with kpeak. We rely on a set of more than a quarter million flux models obtained by varying thermal, reionization histories and cosmological parameters. We find a hint of extra power that is well fitted by the PMF model with B ∼ 0.2 nG, corresponding to kpeak ∼ 20 Mpc−1. However, when applying very conservative assumptions on the modelling of the noise, we obtain a 3σ C.L. lower limit kpeak > 30 Mpc−1 which translates into the tightest bounds on the strength of primordial intergalactic magnetic fields: B © 2025, CC BY.

关键词： Magnetic fields

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：