检索结果-内蒙古大学图书馆

arXiv 2023年

作者： Li, Heng Lin, Ziqin Qiu, Zhongxi Li, Zinan Niu, Ke Guo, Na Fu, Huazhu Hu, Yan Liu, Jiang The Research Institute of Trustworthy Autonomous Systems Southern University of Science and Technology Shenzhen China The Department of Computer Science and Engineering Southern University of Science and Technology Shenzhen China The Institute of High Performance Computing Agency for Science Technology and Research Singapore Computer School Beijing Information Science and Technology University Beijing China The School of Computer and Communication Engineering University of Science and Technology Beijing Beijing China

Medical imaging provides many valuable clues involving anatomical structure and pathological characteristics. However, image degradation is a common issue in clinical practice, which can adversely impact the observation and diagnosis by physicians and algorithms. Although extensive enhancement models have been developed, these models require a well pre-training before deployment, while failing to take advantage of the potential value of inference data after deployment. In this paper, we raise an algorithm for source-free unsupervised domain adaptive medical image enhancement (SAME), which adapts and optimizes enhancement models using test data in the inference phase. A structure-preserving enhancement network is first constructed to learn a robust source model from synthesized training data. Then a teacher-student model is initialized with the source model and conducts source-free unsupervised domain adaptation (SFUDA) by knowledge distillation with the test data. Additionally, a pseudo-label picker is developed to boost the knowledge distillation of enhancement tasks. Experiments were implemented on ten datasets from three medical image modalities to validate the advantage of the proposed algorithm, and setting analysis and ablation studies were also carried out to interpret the effectiveness of SAME. The remarkable enhancement performance and benefits for downstream tasks demonstrate the potential and generalizability of SAME. The code is available at https://***/liamheng/Annotation-free-Medical-Image-Enhancement. Copyright © 2023, The Authors. All rights reserved.

关键词： Distillation

来源：评论

学校读者我要写书评

暂无评论

SMART: towards pre-trained missing-aware model for patient health status prediction 24

SMART: towards pre-trained missing-aware model for patient h...

引用

Proceedings of the 38th International Conference on Neural information Processing Systems

作者： Zhihao Yu Xu Chu Yujie Jin Yasha Wang Junfeng Zhao School of Computer Science Peking University and National Research and Engineering Center of Software Engineering Peking University School of Computer Science Peking University and Center on Frontiers of Computing Studies Peking University Beijing China and Key Laboratory of High Confidence Software Technologies Ministry of Education Beijing China and Peking University Information Technology Institute (Tianjin Binhai) National Research and Engineering Center of Software Engineering Peking University and Key Laboratory of High Confidence Software Technologies Ministry of Education Beijing China and Peking University Information Technology Institute (Tianjin Binhai) School of Computer Science Peking University and National Research and Engineering Center of Software Engineering Peking University and Key Laboratory of High Confidence Software Technologies Ministry of Education Beijing China and Nanhu Laboratory Jiaxing China

ISBN: (纸本)9798331314385

Electronic health record (EHR) data has emerged as a valuable resource for analyzing patient health status. However, the prevalence of missing data in EHR poses significant challenges to existing methods, leading to spurious correlations and suboptimal predictions. While various imputation techniques have been developed to address this issue, they often obsess difficult-to-interpolate details and may introduce additional noise when making clinical predictions. To tackle this problem, we propose SMART, a Self-Supervised Missing-Aware RepresenTation Learning approach for patient health status prediction, which encodes missing information via missing-aware temporal and variable attentions and learns to impute missing values through a novel self-supervised pre-training approach which reconstructs missing data representations in the latent space rather than in input space as usual. By adopting elaborated attentions and focusing on learning higher-order representations, SMART promotes better generalization and robustness to missing data. We validate the effectiveness of SMART through extensive experiments on six EHR tasks, demonstrating its superiority over state-of-the-art methods. Our code is available at https://***/yzhHoward/SMART.

关键词：

来源：评论

学校读者我要写书评

暂无评论

SEDNet: Synergistic Learning Network with Embedded Encoder and Dense Atrous Convolution for Vehicle Re-identification

引用

Alexandria engineering Journal 2025年 128卷 297-305页

作者： Mingfu Xiong Tanghao Gui Zhihong Sun Saeed Anwar Aziz Alotaibi Khan Muhammad School of Computer Science and Artificial Intelligence Wuhan Textile University Wuhan 430200 China Department of Information Security Naval University of Engineering Wuhan 430030 China The University of Western Australia (UWA) Crawley 6009 Australia Department of Computer Science College of Computing and Information Technology Taif University Taif 21974 Saudi Arabia Visual Analytics for Knowledge Laboratory (VIS2KNOW Lab) Department of Applied Artificial Intelligence College of Computing and Informatics Sungkyunkwan University Seoul 03063 South Korea

To address the issue of information redundancy (such as color and vehicle model) caused by excessive emphasis on local features in vehicle re-identification, this paper proposes a Synergistic Learning Network with Embedded Encoder and Dense Atrous Convolution (SEDNet). The proposed SEDNet framework consists of three unique branches: a global embedded multi-head encoder (GEME), local dual-dense atrous convolution (LDAC), and auxiliary attribute embedding (AAM). The GEME branch integrates the global appearance features of the vehicle to enhance consistency in descriptions from different perspectives. To suppress redundant information such as color and vehicle model information, and refine local features, the LDAC branch employs an attention mechanism to capture multiscale features using convolutional kernels with varying dilation rates. In addition, the AAM branch uses vehicle metadata, such as direction and camera perspectives, to enhance feature robustness. Our proposed SEDNet method has been rigorously tested on the mainstream benchmark vehicle re-identification datasets, including VeRi-776, VehicleID, and VeRi-Wild. The results show that our method enhances the mAP by 2.2%, 2.2%, and 0.2%, respectively, when compared to the latest methods, all evaluated on a regular scale. Additional experiments conducted on the Market-1501 and DukeMTMC-reID datasets further verify our method’s generalization capability.

关键词： Vehicle re-identification Embedded encoder Dense atrous convolution Intelligent networks Synergistic learning

来源：评论

学校读者我要写书评

暂无评论

GUEST EDITORIAL

引用

China Communications 2023年第2期 3-6页

作者： Yuan Wu Xu Chen Feng Lyu Xianfu Chen Xumin Huang Jie Gao Yueyue Dai Tony Q.S.Quek Yan Zhang the State Key Laboratory of Internet of Things for Smart City University of Macau the Department of Computer and Information Science University of Macau Sun Yat-sen University Sun Yat-sen University Institute of Advanced Networking and Computing Systems Sun Yat-sen University National and Local Joint Engineering Laboratory the School of Computer Science and Engineering Central South University the VTT Technical Research Centre of Finland the School of Automation Guangdong University of Technology the School of Information Technology Carleton University Huazhong University of Science and Technology Singapore University of Technology and Design(SUTD) the IEEE Communications Society Technical Committee on Green Communications and Computing(TCGCC)

With the growing maturity of the advanced edge-cloud collaboration and integrated sensing-communication-computing systems, edge intelligence has been envisioned as one of the enabling technologies for ubiquitous and l...

关键词：

来源：评论

学校读者我要写书评

暂无评论

VHASR: A Multimodal Speech Recognition System With Vision Hotwords

arXiv

引用

arXiv 2024年

作者： Hu, Jiliang Li, Zuchao Wang, Ping Ai, Haojun Zhang, Lefei Zhao, Hai Key Laboratory of Aerospace Information Security and Trusted Computing Ministry of Education School of Cyber Science and Engineering Wuhan University Wuhan China School of Computer Science Wuhan University Wuhan China School of Information Management Wuhan University Wuhan China Department of Computer Science and Engineering Shanghai Jiao Tong University China

The image-based multimodal automatic speech recognition (ASR) model enhances speech recognition performance by incorporating audio-related image. However, some works suggest that introducing image information to model does not help improving ASR performance. In this paper, we propose a novel approach effectively utilizing audio-related image information and set up VHASR, a multimodal speech recognition system that uses vision as hotwords to strengthen the model’s speech recognition capability. Our system utilizes a dual-stream architecture, which firstly transcribes the text on the two streams separately, and then combines the outputs. We evaluate the proposed model on four datasets: Flickr8k, ADE20k, COCO, and OpenImages. The experimental results show that VHASR can effectively utilize key information in images to enhance the model’s speech recognition ability. Its performance not only surpasses unimodal ASR, but also achieves SOTA among existing image-based multimodal ASR.1 Copyright © 2024, The Authors. All rights reserved.

关键词： Speech recognition

来源：评论

学校读者我要写书评

暂无评论

Efficient learning of continuous-variable quantum states

引用

Physical Review Research 2024年第3期6卷 033280页

作者： Ya-Dong Wu Yan Zhu Giulio Chiribella Nana Liu John Hopcroft Center for Computer Science QICI Quantum Information and Computation Initiative Department of Computer Science The Department of Computer Science for Theoretical Physics Waterloo Ontario N2L 2Y5 Canada Institute of Natural Sciences Ministry of Education Key Laboratory in Scientific and Engineering Computing School of Mathematical Sciences University of Michigan-

The characterization of continuous-variable quantum states is crucial for applications in quantum communication, sensing, simulation, and computing. However, a full characterization of multimode quantum states requires a number of experiments that grows exponentially with the number of modes. Here we propose an alternative approach where the goal is not to reconstruct the full quantum state, but rather to estimate its characteristic function at a given set of points. For multimode states with reflection symmetry, we show that the characteristic function at M points can be estimated using only O(logM) copies of the state, independently of the number of modes. When the characteristic function is known to be positive, as in the case of squeezed vacuum states, the estimation is achieved by an experimentally friendly setup using only beamsplitters and homodyne measurements.

关键词： Machine learning Quantum information processing with continuous variables Quantum tomography

来源：评论

学校读者我要写书评

暂无评论

GLCF: A Global-Local Multimodal Coherence Analysis Framework for Talking Face Generation Detection

arXiv

引用

arXiv 2024年

作者： Chen, Xiaocan Yin, Qilin Liu, Jiarui Lu, Wei Luo, Xiangyang Zhou, Jiantao School of Computer Science and Engineering MoE Key Laboratory of Information Technology Guangdong Province Key Laboratory of Information Security Technology Sun Yat-sen University Guangzhou510006 China Alibaba Group Hangzhou China State Key Laboratory of Mathematical Engineering and Advanced Computing Zhengzhou450002 China State Key Laboratory of Internet of Things for Smart City Department of Computer and Information Science Faculty of Science and Technology University of Macau 999078 China

Talking face generation (TFG) allows for producing lifelike talking videos of any character using only facial images and accompanying text. Abuse of this technology could pose significant risks to society, creating the urgent need for research into corresponding detection methods. However, research in this field has been hindered by the lack of public datasets. In this paper, we construct the first large-scale multi-scenario talking face dataset (MSTF), which contains 22 audio and video forgery techniques, filling the gap of datasets in this field. The dataset covers 11 generation scenarios and more than 20 semantic scenarios, closer to the practical application scenario of TFG. Besides, we also propose a TFG detection framework, which leverages the analysis of both global and local coherence in the multimodal content of TFG videos. Therefore, a region-focused smoothness detection module (RSFDM) and a discrepancy capture-time frame aggregation module (DCTAM) are introduced to evaluate the global temporal coherence of TFG videos, aggregating multi-grained spatial information. Additionally, a visual-audio fusion module (V-AFM) is designed to evaluate audiovisual coherence within a localized temporal perspective. Comprehensive experiments demonstrate the reasonableness and challenges of our datasets, while also indicating the superiority of our proposed method compared to the state-of-the-art deepfake detection approaches. © 2024, CC BY-NC-ND.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

SUMI-IFL: An information-Theoretic Framework for Image Forgery Localization with Sufficiency and Minimality Constraints

arXiv

引用

arXiv 2024年

作者： Sheng, Ziqi Lu, Wei Luo, Xiangyang Zhou, Jiantao Cao, Xiaochun School of Computer Science and Engineering MoE Key Laboratory of Information Technology Guangdong Province Key Laboratory of Information Security Technology Sun Yat-sen University China State Key Laboratory of Mathematical Engineering and Advanced Computing China State Key Laboratory of Internet of Things for Smart City Department of Computer and Information Science University of Macau China School of Cyber Science and Technology Sun Yat-sen University Shenzhen Campus China

Image forgery localization (IFL) is a crucial technique for preventing tampered image misuse and protecting social safety. However, due to the rapid development of image tampering technologies, extracting more comprehensive and accurate forgery clues remains an urgent challenge. To address these challenges, we introduce a novel information-theoretic IFL framework named SUMI-IFL that imposes sufficiency-view and minimality-view constraints on forgery feature representation. First, grounded in the theoretical analysis of mutual information, the sufficiency-view constraint is enforced on the feature extraction network to ensure that the latent forgery feature contains comprehensive forgery clues. Considering that forgery clues obtained from a single aspect alone may be incomplete, we construct the latent forgery feature by integrating several individual forgery features from multiple perspectives. Second, based on the information bottleneck, the minimality-view constraint is imposed on the feature reasoning network to achieve an accurate and concise forgery feature representation that counters the interference of task-unrelated features. Extensive experiments show the superior performance of SUMI-IFL to existing state-of-the-art methods, not only on in-dataset comparisons but also on cross-dataset comparisons. © 2024, CC BY-NC-ND.

关键词： Clutter (information theory)

来源：评论

学校读者我要写书评

暂无评论

A digital 3D TCAM accelerator for the inference phase of Random Forest

A digital 3D TCAM accelerator for the inference phase of Ran...

引用

Design Automation Conference

作者： Chieh-Lin Tsai Chun-Feng Wu Yuan-Hao Chang Han-Wen Hu Yung-Chun Lee Hsiang-Pang Li Tei-Wei Kuo Department of Computer Science and Information Engineering National Taiwan University Department of Computer Science National Yang Ming Chiao Tung University Institute of Information Science Academia Sinica Taiwan Macronix International Co. Ltd Department of Electrical Engineering National Tsing Hua University Mohamed Bin Zayed University of Artificial Intelligence NTU High Performance and Scientific Computing Center National Taiwan University

Random forest is a popular ensemble machine-learning algorithm for classification and regression tasks. However, the irregular tree shapes and non-deterministic memory access patterns make it hard for the current von Neumann architecture to handle random forest efficiently. This paper proposes a digital 3D TCAM-based accelerator for the random forest, adopting the idea of processing-in-memory (PIM) to reduce data movement. By utilizing this accelerator, we propose a TCAM-based approach to provide real-time inference with low energy consumption, making it suitable for edge or embedded environments. In the experiments, the proposed approach achieves an average of 3.13 times higher throughput with 22 times more energy saving than the GPU approach.

关键词：

来源：评论

学校读者我要写书评

暂无评论

On the properties of kullback-leibler divergence between multivariate Gaussian distributions 23

On the properties of kullback-leibler divergence between mul...

引用

Proceedings of the 37th International Conference on Neural information Processing Systems

作者： Yufeng Zhang Jialu Pan Kenli Li Wanwei Liu Zhenbang Chen Xinwang Liu Ji Wang College of Computer Science and Electronic Engineering Hunan University Changsha China College of Computer Key Laboratory of Software Engineering for Complex Systems National University of Defense Technology Changsha China College of Computer State Key Laboratory for High Performance Computing Key laboratory of Software Engineering for Complex Systems National University of Defense Technology Changsha China

Kullback-Leibler (KL) divergence is one of the most important measures to calculate the difference between probability distributions. In this paper, we theoretically study several properties of KL divergence between multivariate Gaussian distributions. Firstly, for any two n-dimensional Gaussian distributions Ɲ1 and Ɲ2, we prove that when KL(Ɲ2‖Ɲ1) ≤ ε (ε > 0) the supremum of KL (Ɲ1‖Ɲ2) is (1/2) ((-W0 (-e-(1+2ε)))-1 + log(-W0 (-e-(1+2ε))) - 1), where W0 is the principal branch of Lambert W function. For small ε, the supremum is ε + 2ε1.5 + O (ε2). This quantifies the approximate symmetry of small KL divergence between Gaussian distributions. We further derive the infimum of KL(Ɲ1‖Ɲ2) when KL(Ɲ2‖Ɲ1) ≥ M (M > 0). We give the conditions when the supremum and infimum can be attained. Secondly, for any three n-dimensional Gaussian distributions Ɲ1, Ɲ2, and Ɲ3, we theoretically show that an upper bound of KL (Ɲ1‖Ɲ3) is 3ε1 + 3ε2 + 2√ε1ε2 + o(ε1)+ o(ε2) when KL (Ɲ1‖Ɲ2) ≤ ε1 and KL(Ɲ2‖Ɲ3) ≤ ε2 (ε1, ε2 ≥ 0). This reveals that KL divergence between Gaussian distributions follows a relaxed triangle inequality. Note that, all these bounds in the theorems presented in this work are independent of the dimension n. Finally, we discuss several applications of our theories in deep learning, reinforcement learning, and sample complexity research.

关键词：

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：