检索结果-内蒙古大学图书馆

arXiv 2023年

作者： Gochoo, Munkhjargal Otgonbold, Munkh-Erdene Ganbold, Erkhembayar Hsieh, Jun-Wei Chang, Ming-Ching Chen, Ping-Yang Dorj, Byambaa Al Jassmi, Hamad Batnasan, Ganzorig Alnajjar, Fady Abduljabbar, Mohammed Lin, Fang-Pang Department of Computer Science and Software Engineering United Arab Emirates University United Arab Emirates Emirates Center for Mobility Research United Arab Emirates University United Arab Emirates College of AI and Green Energy National Yang Ming Chiao Tung University Taiwan University at Albany State University of New York NY United States Department of Computer Science National Yang Ming Chiao Tung University Taiwan Mongolian University of Science and Technology Mongolia National Center for High-Performance Computing Taiwan

With the advance of AI, road object detection has been a prominent topic in computer vision, mostly using perspective cameras. Fisheye lens provides omnidirectional wide coverage for using fewer cameras to monitor road intersections, however with view distortions. To our knowledge, there is no existing open dataset prepared for traffic surveillance on fisheye cameras. This paper introduces an open FishEye8K benchmark dataset for road object detection tasks, which comprises 157K bounding boxes across five classes (Pedestrian, Bike, Car, Bus, and Truck). In addition, we present benchmark results of State-of-The-Art (SoTA) models, including variations of YOLOv5, YOLOR, YOLO7, and YOLOv8. The dataset comprises 8,000 images recorded in 22 videos using 18 fisheye cameras for traffic monitoring in Hsinchu, Taiwan, at resolutions of 1080×1080 and 1280×1280. The data annotation and validation process were arduous and time-consuming, due to the ultra-wide panoramic and hemispherical fisheye camera images with large distortion and numerous road participants, particularly people riding scooters. To avoid bias, frames from a particular camera were assigned to either the training or test sets, maintaining a ratio of about 70:30 for both the number of images and bounding boxes in each class. Experimental results show that YOLOv8 and YOLOR outperform on input sizes 640×640 and 1280×1280, respectively. The dataset will be available on the GitHub link with PASCAL VOC, MS COCO, and YOLO annotation formats. The Fish-Eye8K benchmark will provide significant contributions to the fisheye video analytics and smart city applications. Copyright © 2023, The Authors. All rights reserved.

关键词： Object detection

来源：评论

学校读者我要写书评

暂无评论

On the Equity of Nuclear Norm Maximization in Unsupervised Domain Adaptation

arXiv

引用

arXiv 2022年

作者： Zhang, Wenju Zhang, Xiang Liao, Qing Lan, Long Wang, Mengzhu Wang, Wei Peng, Baoyun Ding, Zhengming Institute for Quantum Information State Key Laboratory of High Performance Computing National University of Defense Technology Changsha China College of Computer National University of Defense Technology Changsha China Department of Computer Science and Technology Harbin Institute of Technology Shenzhen China DUT-RU International School of Information Science & Engineering Dalian University of Technology Dalian China Tulane University New OrleansLA United States

Nuclear norm maximization has shown the power to enhance the transferability of unsupervised domain adaptation model (UDA) in an empirical scheme. In this paper, we identify a new property termed equity, which indicates the balance degree of predicted classes, to demystify the efficacy of nuclear norm maximization for UDA theoretically. With this in mind, we offer a new discriminability-and-equity maximization paradigm built on squares loss, such that predictions are equalized explicitly. To verify its feasibility and flexibility, two new losses termed Class Weighted Squares Maximization (CWSM) and Normalized Squares Maximization (NSM), are proposed to maximize both predictive discriminability and equity, from the class level and the sample level, respectively. Importantly, we theoretically relate these two novel losses (i.e., CWSM and NSM) to the equity maximization under mild conditions, and empirically suggest the importance of the predictive equity in UDA. Moreover, it is very efficient to realize the equity constraints in both losses. Experiments of cross-domain image classification on three popular benchmark datasets show that both CWSM and NSM contribute to outperforming the corresponding counterparts. © 2022, CC BY.

关键词： Image classification

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

arXiv

引用

arXiv 2021年

作者： Zhang, Yufeng Liu, Wanwei Chen, Zhenbang Wang, Ji Li, Kenli The College of Computer Science and Electronic Engineering Hunan University Changsha China The College of Computer National University of Defense Technology Changsha China State Key Laboratory of High Performance Computing National University of Defense Technology China

Kullback-Leibler (KL) divergence is one of the most important divergence measures between probability distributions. In this paper, we prove several properties of KL divergence between multivariate Gaussian distributions. First, for any two n-dimensional Gaussian distributions N1 and N2, we give the supremum of KL(N1||N2) when KL(N2||N1) ≤ Ε (Ε > 0). For small Ε, we show that the supremum is Ε + 2Ε1.5 + O(Ε2). This quantifies the approximate symmetry of small KL divergence between Gaussians. We also find the infimum of KL(N1||N2) when KL(N2||N1) ≥ M (M > 0). We give the conditions when the supremum and infimum can be attained. Second, for any three n-dimensional Gaussians N1, N2, and N3, we find an upper bound of KL(N1||N3) if KL(N1||N2) ≤ Ε1 and KL(N2||N3) ≤ Ε2 for Ε1, Ε2 ≥ 0. For small Ε1 and Ε2, we show the upper bound is 3Ε1 + 3Ε2 + 2√Ε1Ε2 + o(Ε1) + o(Ε2). This reveals that KL divergence between Gaussians follows a relaxed triangle inequality. Importantly, all the bounds in the theorems presented in this paper are independent of the dimension n. Finally, We discuss the applications of our theorems in explaining counterintuitive phenomenon of flow-based model, deriving deep anomaly detection algorithm, and extending one-step robustness guarantee to multiple steps in safe reinforcement learning. Copyright © 2021, The Authors. All rights reserved.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Analysis of Video Quality Datasets via Design of Minimalistic Video Quality Models

arXiv

引用

arXiv 2023年

作者： Sun, Wei Wen, Wen Min, Xiongkuo Lan, Long Zhai, Guangtao Ma, Kede The Institute of Image Communication and Information Processing Shanghai Jiao Tong University Shanghai200240 China The Department of Computer Science City University of Hong Kong Kowloon Hong Kong The Institute for Quantum Information State Key Laboratory of High Performance Computing College of Computer Science and Technology National University of Defense Technology Changsha410073 China The Institute of Image Communication and Information Processing The MoE Key Lab of Artificial Intelligence AI Institute Shanghai Jiao Tong University Shanghai200240 China The Department of Computer Science The Shenzhen Research Institute City University of Hong Kong Kowloon Hong Kong

Blind video quality assessment (BVQA) plays an indispensable role in monitoring and improving the end-users’ viewing experience in various real-world video-enabled media applications. As an experimental field, the improvements of BVQA models have been measured primarily on a few human-rated VQA datasets. Thus, it is crucial to gain a better understanding of existing VQA datasets in order to properly evaluate the current progress in BVQA. Towards this goal, we conduct a first-of-its-kind computational analysis of VQA datasets via designing minimalistic BVQA models. By minimalistic, we restrict our family of BVQA models to build only upon basic blocks: a video preprocessor (for aggressive spatiotemporal downsampling), a spatial quality analyzer, an optional temporal quality analyzer, and a quality regressor, all with the simplest possible instantiations. By comparing the quality prediction performance of different model variants on eight VQA datasets with realistic distortions, we find that nearly all datasets suffer from the easy dataset problem of varying severity, some of which even admit blind image quality assessment (BIQA) solutions. We additionally justify our claims by comparing our model generalization capabilities on these VQA datasets, and by ablating a dizzying set of BVQA design choices related to the basic building blocks. Our results cast doubt on the current progress in BVQA, and meanwhile shed light on good practices of constructing next-generation VQA datasets and models. Code is available at https://***/sunwei925/***. Copyright © 2023, The Authors. All rights reserved.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

Human-cyber-physical systems:concepts,challenges,and research opportunities

引用

Frontiers of Information Technology & Electronic engineering 2020年第11期21卷 1535-1553页

作者： Zhiming LIU Ji WANG RISE-Centre for Research and Innovation in Software Engineering School of Computer and Information ScienceSouthwest UniversityChongqing 400715China State Key Laboratory for High Performance Computing School of ComputerNational University of Defense TechnologyChangsha 410073China

In this perspective article,we first recall the historic background of human-cyber-physical systems(HCPSs),and then introduce and clarify important *** discuss the key challenges in establishing the scientific foundation from a system engineering point of view,including(1)complex heterogeneity,(2)lack of appropriate abstractions,(3)dynamic black-box integration of heterogeneous systems,(4)complex requirements for functionalities,performance,and quality of services,and(5)design,implementation,and maintenance of HCPS to meet *** we propose four research directions to tackle the challenges,including(1)abstractions and computational theory of HCPS,(2)theories and methods of HCPS architecture modelling,(3)specification and verification of model properties,and(4)software-defined *** article also serves as the editorial of this special section on cyber-physical systems and summarises the four articles included in this special section.

关键词： Abstractions Architecture modelling Evolution Software-defined technology

来源：评论

学校读者我要写书评

暂无评论

Loophole-free test of macroscopic realism via high-order correlations of measurement

arXiv

引用

arXiv 2024年

作者： Wang, Ping Chen, Chong Liao, Hao Vorobyov, Vadim V. Wrachtrup, Jörg Liu, Ren-Bao College of Education for the Future Beijing Normal University Zhuhai519087 China Department of Physics The Chinese University of Hong Kong New Territories Shatin Hong Kong Centre for Quantum Coherence The Chinese University of Hong Kong New Territories Shatin Hong Kong The Hong Kong Institute of Quantum Information Science and Technology The Chinese University of Hong Kong New Territories Shatin Hong Kong National Engineering Laboratory on Big Data System Computing Technology Guangdong Province Engineering Center of China-made High Performance Data Computing System College of Computer Science and Software Engineering Shenzhen University Shenzhen518060 China 3rd Institute of Physics Research Center SCoPE and IQST University of Stuttgart Stuttgart70569 Germany New Cornerstone Science Laboratory The Chinese University of Hong Kong New Territories Shatin Hong Kong

Test of macroscopic realism (MR) is key to understanding the foundation of quantum mechanics. Due to the existence of the non-invasive measurability loophole and other interpretation loopholes, however, such test remains an open question. Here we propose a general inequality based on high-order correlations of measurements for a loophole-free test of MR at the weak signal limit. Importantly, the inequality is established using the statistics of raw data recorded by classical devices, without requiring a specific model for the measurement process, so its violation would falsify MR without the interpretation loophole. The non-invasive measurability loophole is also closed, since the weak signal limit can be verified solely by measurement data (using the relative scaling behaviors of different orders of correlations). We demonstrate that the inequality can be broken by a quantum spin model. The inequality proposed here provides an unambiguous test of the MR principle and is also useful to characterizing quantum coherence. Copyright © 2024, The Authors. All rights reserved.

关键词： Quantum theory

来源：评论

学校读者我要写书评

暂无评论

Statistical Modeling of Soft Error Influence on Neural Networks

arXiv

引用

arXiv 2022年

作者： Huang, Haitong Xue, Xinghua Liu, Cheng Wang, Ying Luo, Tao Cheng, Long Li, Huawei Li, Xiaowei Beijing100190 China Department of Computer Science University of Chinese Academy of Sciences Beijing100190 China Beijing100190 China Peng Cheng Laboratory Shenzhen518055 China Institute of High Performance Computing A*STAR 138632 Singapore North China Electric Power University Beijing102206 China

Soft errors in large VLSI circuits pose dramatic influence on computing- and memory-intensive neural network (NN) processing. Understanding the influence of soft errors on NNs is critical to protect against soft errors for reliable NN processing. Prior work mainly rely on fault simulation to analyze the influence of soft errors on NN processing. They are accurate but usually specific to limited configurations of errors and NN models due to the prohibitively slow simulation speed especially for large NN models and datasets. With the observation that the influence of soft errors propagates across a large number of neurons and accumulates as well, we propose to characterize the soft error induced data disturbance on each neuron with normal distribution model according to central limit theorem and develop a series of statistical models to analyze the behavior of NN models under soft errors in general. The statistical models reveal not only the correlation between soft errors and NN model accuracy, but also how NN parameters such as quantization and architecture affect the reliability of NNs. The proposed models are compared with fault simulation and verified comprehensively. In addition, we observe that the statistical models that characterize the soft error influence can also be utilized to predict fault simulation results in many cases and we explore the use of the proposed statistical models to accelerate fault simulations of NNs. According to our experiments, the accelerated fault simulation shows almost two orders of magnitude speedup with negligible simulation accuracy loss over the baseline fault simulations. Copyright © 2022, The Authors. All rights reserved.

关键词： Neural network models

来源：评论

学校读者我要写书评

暂无评论

FedCache: A Knowledge Cache-driven Federated Learning Architecture for Personalized Edge Intelligence

arXiv

引用

arXiv 2023年

作者： Wu, Zhiyuan Sun, Sheng Wang, Yuwei Liu, Min Xu, Ke Wang, Wen Jiang, Xuefeng Gao, Bo Lu, Jinda The Institute of Computing Technology Chinese Academy of Sciences Beijing China The University of Chinese Academy of Sciences Beijing China The Zhongguancun Laboratory Beijing China The Department of Computer Science and Technology Tsinghua University Beijing China The School of Computer and Information Technology The Engineering Research Center of Network Management Technology for High-Speed Railway Ministry of Education Beijing Jiaotong University Beijing China The School of Information Science and Technology University of Science and Technology of China Hefei China

Edge Intelligence (EI) allows Artificial Intelligence (AI) applications to run at the edge, where data analysis and decision-making can be performed in real-time and close to data sources. To protect data privacy and unify data silos distributed among end devices in EI, Federated Learning (FL) is proposed for collaborative training of shared AI models across multiple devices without compromising data privacy. However, the prevailing FL approaches cannot guarantee model generalization and adaptation on heterogeneous clients. Recently, Personalized Federated Learning (PFL) has drawn growing awareness in EI, as it enables a productive balance between local-specific training requirements inherent in devices and global-generalized optimization objectives for satisfactory performance. However, most existing PFL methods are based on the Parameters Interaction-based Architecture (PIA) represented by FedAvg, which suffers from unaffordable communication burdens due to large-scale parameters transmission between devices and the edge server. In contrast, Logits Interaction-based Architecture (LIA) allows to update model parameters with logits transfer and gains the advantages of communication lightweight and heterogeneous on-device model allowance compared to PIA. Nevertheless, previous LIA methods attempt to achieve satisfactory performance either relying on unrealistic public datasets or increasing communication overhead for additional information transmission other than logits. To tackle this dilemma, we propose a knowledge cache-driven PFL architecture, named FedCache, which reserves a knowledge cache on the server for fetching personalized knowledge from the samples with similar hashes to each given on-device sample. During the training phase, ensemble distillation is applied to on-device models for constructive optimization with personalized knowledge transferred from the server-side knowledge cache. Empirical experiments on four datasets demonstrate that FedCache achiev

关键词： Distillation

来源：评论

学校读者我要写书评

暂无评论

A Novel Deep Learning Approach Featuring Graph-Based Algorithm for Cell Segmentation and Tracking

A Novel Deep Learning Approach Featuring Graph-Based Algorit...

引用

IEEE International Conference on Bioinformatics and Biomedicine (BIBM)

作者： Keliang Zhao Jovial Niyogisubizo Linxia Xiao Yi Pan Dongqing Wei Didi Rosiyadi Yanjie Wei Shenzhen Key Laboratory of Intelligent Bioinformatics and Center for High Performance Computing Shenzhen Institute of Advanced Technology Chinese Academy of Sciences Shenzhen China College of Computer Science and Control Engineering Shenzhen Institutes of Advanced Technology Chinese Academy of Sciences Shenzhen China State Key Laboratory of Microbial Metabolism and School of Life Sciences and Biotechnology Shanghai Jiao Tong University Shanghai China Research Center for Artificial Intelligence and Cybersecurity National Research and Innovation Agency Bandung Indonesia

The precise segmentation and tracking of cells in microscopy image sequences play a pivotal role in biomedical research, facilitating the study of tissue, organ, and organism development. However, manual segmentation and tracking of cells is time-consuming and often require professional experiences. Besides, segmenting cells in the images with a low signal-to-noise ratio remains difficult. While deep learning (DL) has become a common method for cell segmentation, few DL- based methods address concurrent cell segmentation and tracking. In this paper, we propose a novel DL approach featuring graph-based tracking for cell segmentation and tracking in microscopy images. We combine Deeplabv3+ for semantic segmentation and ResNet50 for enhanced feature extraction, enabling comprehensive cell detection and instance segmentation. Post-processing, involving non-maxima suppression and outlier detection, refines predictions and produces final segmentation. The tracking method is based on the relative position of graph nodes to track segmented cells, encompassing cell division and apoptosis. We conduct our experiments on the induced pluripotent stem (iPS) cell datasets, and the results show that the segmentation and tracking performance of our method yields superior performance compared to the benchmark models. More specifically, our approach achieved DET values of 0.955 and 0.913, TRA values of 0.951 and 0.906, and SEG values of 0.690 and 0.665 on two iPS dataset videos, respectively. Additionally, the performance assessment encompasses four real microscopy datasets from the Cell Tracking Challenge (CTC). Our method greatly reduces the cost of manual labeling and labor-intensive costs.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Hsm2: A hybrid and scalable metadata management method in distributed file systems 10th

Hsm2: A hybrid and scalable metadata management method in di...

引用

10th International Symposium on Parallel Architectures, Algorithms and Programming, PAAP 2019

作者： Wang, Yiduo Chen, Youxu Shao, Xinyang Chen, Jinzhong Yuan, Liu Xu, Yinlong School of Computer Science and Technology University of Science and Technology of China AnHui Province Key Laboratory of High Performance Computing Hefei China East China Research Institute of Electronic Engineering Hefei China China Academy of Electronics and Information Technology Beijing China

In the bigdata era, metadata performance is critical in modern distributed file systems. Traditionally, the metadata management strategies like the subtree partitioning method focus on keeping namespace locality, while the other ones like the hash-based mapping method aim to offer good load balance. Nevertheless, none of these methods achieve the two desirable properties simultaneously. To close this gap, in this paper, we propose a novel metadata management scheme, HSM2, which combines the subtree partitioning and hash-based mapping method together. We implemented HSM2 in CephFS, a widely deployed distributed file systems, and conducted a comprehensive set of metadata-intensive experiments. Experimental results show that HSM2 can achieve better namespace locality and load balance simultaneously. Compared with CephFS, HSM2 can reduce the completion time by 70% and achieve 3.9× overall throughput speedup for a file-scanning workload. © Springer Nature Singapore Pte Ltd. 2020.

关键词： Metadata

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：