检索结果-内蒙古大学图书馆

SeaFree-GS: Reconstructing Underwater 3D Scenes With True Appearances

IEEE Signal Processing Letters 2025年 32卷 2114-2118页

作者： Liu, Shaohua Gao, Ning Fu, Shaowen Zhong, Xiaoqing Li, Hongjue Beihang University School of Astronautics Shenyuan Honors College Beijing100191 China Beihang University School of Astronautics Beijing100191 China Beijing Aerospace Automatic Control Institute Beijing100854 China National Key Laboratory of Science and Technology on Aerospace Intelligent Control Beijing100854 China China Academy of Space Technology Beijing100094 China Beijing Key Laboratory of System Design for Reusable Launch Vehicle State Key Laboratory of High- Efficiency Reusable Aerospace Transportation Technology China

Reconstructing underwater 3D scenes with accurate appearances is crucial for numerous tasks. However, existing underwater 3D reconstruction methods often fail to restore the true scene appearance due to degradation in underwater images caused by water effects. In this letter, we propose SeaFree-GS, a novel approach leveraging 3D Gaussian Splatting (3DGS) to reconstruct underwater scenes with their true appearances. Specifically, we introduce a Degradation-Aware Dual-Color Modeling strategy, where each Gaussian is assigned an intrinsic color representing the true scene appearance and a viewpoint-dependent degraded color to incorporate water effects. For a given viewpoint, this strategy physically derives the corresponding degraded colors from the intrinsic colors to render the underwater image. To improve reconstruction accuracy, we introduce a Content-Based Loss for selective enhancement of supervision over foreground and background regions, and a Coarse-Grained Depth Loss to enforce additional geometric constraints. Experiments on three datasets demonstrate that SeaFree-GS achieves state-of-the-art performance in Underwater True Appearance Reconstruction, and also performs competitively in Underwater Novel View Synthesis. © 1994-2012 IEEE.

关键词： 3D reconstruction

来源：评论

学校读者我要写书评

暂无评论

Chirp-free hundred-picosecond pulse fiber laser based on carbon nanotubes mode-locker

引用

Optics Express 2025年第7期33卷 14850-14857页

作者： Li, Weixi Jiang, Kailin Dai, Lilong Huang, Qianqian Zou, Meng Luo, Siyu Chen, Siyu Song, Yingxiong Liu, Yunqi Yan, Zhijun Mou, Chengbo Key Laboratory of Specialty Fiber Optics and Optical Access Networks Shanghai Institute for Advanced Communication and Data Science Joint International Research Laboratory of Specialty Fiber Optics and Advanced Communication Shanghai University Shanghai200444 China School of Optical and Electronic Information National Engineering Laboratory for Next Generation Internet Access System Huazhong University of Science and Technology Wuhan430074 China

We demonstrate a narrow bandwidth passively erbium-doped mode-locked fiber laser based on single-wall carbon nanotubes (SWCNTs). The center wavelength is 1546.61 nm. The strong filtering effect and low saturable absorber (SA) modulation depth provide advantages for a long pulse duration of 323 ps. An overall spectral bandwidth of 1.3 GHz (0.01 nm) is measured by the optical heterodyne method, representing the narrowest spectral width based on SWCNT under a single filter structure. The obtained spectral width is very close to the transform-limited bandwidth of 8 pm, indicating that the pulse is chirp-free. Meanwhile, the generation of beat signals reveals the strong coherence between pulses. The repetition frequency of the laser is 7.8 MHz, and the number of longitudinal mode locks is only 166, which is the minimum number of longitudinal mode-locking in passively mode-locked laser based on material SA. At the same time, numerical simulations further support our experimental results and reveal the pulse evolution, providing evidence that SA and filter are the key factors in the generation of chirp-free long pulse. © 2025 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement.

关键词： Saturable absorbers

来源：评论

学校读者我要写书评

暂无评论

Towards Stricter Black-box Integrity Verification of Deep Neural Network Models 24

Towards Stricter Black-box Integrity Verification of Deep Ne...

引用

32nd ACM International Conference on Multimedia, MM 2024

作者： He, Chaoxiang Bai, Xiaofan Ma, Xiaojing Zhu, Bin B. Hu, Pingyi Fu, Jiayun Jin, Hai Zhang, Dongmei Huazhong University of Science and Technology Hubei Wuhan China Shanghai Jiao Tong University Shanghai China Microsoft Corporation Beijing China National Engineering Research Center for Big Data Technology and System Services Computing Technology and System Lab Hubei Engineering Research Center on Big Data Security Hubei Key Laboratory of Distributed System Security School of Cyber Science and Engineering Huazhong University of Science and Technology 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 China

ISBN: (纸本)9798400706868

Cloud-based machine learning services offer significant advantages but also introduce the risk of tampering with cloud-deployed deep neural network (DNN) models. Black-box integrity verification (BIV) allows model owners and end-users to determine if a cloud-deployed DNN model has been tampered with by examining only the top-1 label responses. Fingerprinting generates fingerprint samples to query the model, achieving BIV with no impact on the model's accuracy. In this paper, we present BIVBench, the first comprehensive benchmark for BIV of DNN models. BIVBench covers 16 types of model modifications, providing extensive coverage of practical modification scenarios. Our analysis reveals that existing fingerprinting methods, which are typically focused on significant tampering, lack the sensitivity needed to effectively detect subtle yet common and potentially severe modifications. To address this limitation, we propose MiSentry (Model Integrity Sentry), a novel fingerprinting method that leverages meta-learning. MiSentry strategically incorporates a few subtly modified models into the meta-learning model zoo and maximizes the divergence of output predictions between the target model and the modified models in the model zoo to generate highly sensitive, generalizable, and effective fingerprint samples. Extensive evaluations using BIVBench demonstrate that MiSentry outperforms existing state-of-the-art methods overall and significantly surpasses them in detecting subtle modifications. The BIVBench and supplementary materials are available at: https://***/CGCL-codes/BIVBench. © 2024 ACM.

关键词： Deep neural networks

来源：评论

学校读者我要写书评

暂无评论

Research on Image Captioning Based on Vision-language Pre-trained Models 9

Research on Image Captioning Based on Vision-language Pre-tr...

引用

9th International Conference on Big data and Information Analytics, BigDIA 2023

作者： Zou, Shiwei Xie, Yuxiang Yan, Jie Wei, Yingmei Luan, Xidao National University of Defense Technology Department of System Engineering Hunan China National University of Defense Technology Key Laboratory of Big Data Department of System Engineering Changsha Hunan China Changsha University Department of Computer Science Hunan China

ISBN: (纸本)9798350330076

Image captioning is a vision-language task that targets at describing an image by generating a coherent sentence automatically. This technology allows computers to understand and describe images like humans, enabling further processing of images. It combines computer vision and natural language processing, which has broad applications. This paper investigates the use of vision-language pre-trained models for image captioning. In this article, an encoder-decoder hybrid model based on PVT and BERT is pre-trained on a large number of image-text pairs, possessing both understanding and generation capabilities. It consists of three components: 1) an image encoder and a text encoder to extract image and text features respectively;2) a multi-modal encoder to align the two modalities;and 3) a text decoder to generate text descriptions. This paper focuses on enhancing the image encoder of the model, which is crucial for image captioning models as it extracts high-resolution image features. Reducing training time and computational costs can also improve the applicability of pre-trained models. Improvements are made to the original model by modifying the ViT model architecture of the image encoding module. A feature pyramid structure and a spatial reduction attention mechanism is introduced to extract high-resolution image features while reducing computational complexity and memory usage, thereby enhancing the universality of the model. The model adopts a pure Transformer architecture, discarding the convolutional structure, and achieves excellent performance. © 2023 IEEE.

关键词： Image enhancement

来源：评论

学校读者我要写书评

暂无评论

EagerLog: Active Learning Enhanced Retrieval Augmented Generation for Log-based Anomaly Detection

EagerLog: Active Learning Enhanced Retrieval Augmented Gener...

引用

2025 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2025

作者： Duan, Chiming Jia, Tong Yang, Yong Liu, Guiyang Liu, Jinbu Zhang, Huxing Zhou, Qi Li, Ying Huang, Gang School of Software and Microelectronics Peking University Beijing China Institute for Artificial Intelligence Peking University Beijing China National Engineering Research Center for Software Engineering Peking University Beijing China National Key Laboratory of Data Space Technology and System Beijing China Alibaba Cloud Computing Company Hangzhou China

ISBN: (纸本)9798350368741

Logs record essential information about system operations and serve as a critical source for anomaly detection, which has generated growing research interest. Utilizing large language models (LLMs) within a retrieval-augmented generation (RAG) framework for log-based anomaly detection is an effective approach due to its strong generalization capabilities and efficient few-shot performance. However, the effectiveness of this method hinges on the quality of the knowledge source, which can be impacted by noise and changes within the software systems. Facing these problems, in this paper, we propose a novel log-based anomaly detection method named EagerLog, employing active learning to choose the logs for humans to label, thereby adding them to the knowledge source, thus enhancing the knowledge source and maintaining its quality. Our experiments on three open datasets (BGL, Thunderbird, Zookeeper) and one industrial dataset demonstrate that EagerLog can achieve 93.65% F1 score with approximately 10 labeled log sequences, surpassing existing methods by 15.32%. © 2025 IEEE.

关键词： Active Learning Anomaly Detection Log Analysis Retrieval Augmented Generation

来源：评论

学校读者我要写书评

暂无评论

Differentially Private and Heterogeneity-Robust Federated Learning With Theoretical Guarantee

IEEE Transactions on Artificial Intelligence

引用

IEEE Transactions on Artificial Intelligence 2024年第12期5卷 6369-6384页

作者： Wang, Xiuhua Wang, Shuai Li, Yiwei Fan, Fengrui Li, Shikang Lin, Xiaodong Huazhong University of Science and Technology Hubei Key Laboratory of Distributed System Security Hubei Engineering Research Center on Big Data Security School of Cyber Science and Engineering Wuhan430074 China University of Electronic Science and Technology of China National Key Laboratory of Wireless Communications Chengdu611731 China Xiamen University of Technology Fujian Key Laboratory of Communication Network and Information Processing Xiamen361024 China University of Guelph School of Computer Science GuelphONN1G 2W1 Canada

Federated learning (FL) is a popular distributed paradigm where enormous clients collaboratively train a machine learning (ML) model under the orchestration of a central server without knowing the clients' private raw data. The development of effective FL algorithms faces multiple practical challenges including data heterogeneity and clients' privacy protection. Despite that numerous attempts have been made to deal with data heterogeneity or rigorous privacy protection, none have effectively tackled both issues simultaneously. In this article, we propose a differentially private and heterogeneity-robust FL algorithm, named DP-FedCVR to mitigate the data heterogeneity by following the client-variance-reduction strategy. Besides, it adopts a sophisticated differential privacy (DP) mechanism where the privacy-amplified strategy is applied, to achieve a rigorous privacy protection guarantee. We show that the proposed DP-FedCVR algorithm maintains its heterogeneity-robustness though DP noises are incorporated, while achieving a sublinear convergence rate for a nonconvex FL problem. Numerical experiments based on image classification tasks are presented to demonstrate that DP-FedCVR provides superior performance over the benchmark algorithms in the presence of data heterogeneity and various DP privacy budgets. © 2020 IEEE.

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

TERMS: Task management policies to achieve high performance for mixed workloads using surplus resources

引用

Journal of Parallel and Distributed Computing 2022年第0期170卷 74-85页

作者： Yu, Jinyu Tong, Wei Lv, Pengze Feng, Dan Wuhan National Laboratory for Optoelectronics Key Laboratory of Information Storage System Engineering Research Center of Data Storage Systems and Technology School of Computer Science and Technology Huazhong University of Science and Technology Ministry of Education of China Hubei Wuhan China

Resource contentions and performance interferences can lead to workload performance degradation in mixed-workload deployment clusters. Previous work guarantees the resource requirements of latency-sensitive tasks and reduces performance losses to batch jobs by reclaiming surplus resources from over-provisioned tasks. While the fragmentation of resources leads to a mismatch between provisioned resources and task requirements, resulting in high operation overheads and losses of task fairness. This paper proposes TERMS, the task management policies based on task relevance, resource distribution, and task fairness to achieve efficient and low-cost task management. TERMS mainly includes three types of management policies. The task scheduling policy can schedule new tasks according to task relevance. Task selection strategies select tasks for resource provisioning and task resumption based on resource requirements and task fairness. If necessary, the node selection strategy can be used to choose befitting target nodes based on task relevance and node resource information for task migration when eliminating straggler tasks. Evaluation results show that TERMS can further improve the performance of latency-sensitive services and batch jobs, reduce management overheads, and avoid operation failures. © 2022 Elsevier Inc.

关键词： Scheduling

来源：评论

学校读者我要写书评

暂无评论

Water storage changes in the North China Plain from 2004 to 2019

引用

Science China(Technological Sciences) 2023年第7期66卷 2036-2046页

作者： CHEN MinLan XU YuYue PAN Hang WANG LuHan School of Geography and Ocean Science Nanjing UniversityNanjing 210023China Jiangsu Provincial Key Laboratory of Geographic Information Science and Technology Nanjing UniversityNanjing 210023China Frontiers Science Center for Critical Earth Material Cycling Nanjing UniversityNanjing 210023China National Earth System Science Data Center National Science and Technology Infrastructure of ChinaBeijing 100101China

In the past few decades,the irrational use of water resources has resulted in many issues such as land subsidence in the North China Plain(NCP),hindering its socio-economic *** accurate understanding of water resource changes is important for the allocation of water resources in the *** this study,we employed Gravity Recovery and Climate Experiment(GRACE) satellite data to monitor the total water storage(TWS) change in the NCP during 2004–***,precipitation,and runoff during 2004–2019 were sequentially examined,using the water balance formula(WBF),to retrieve TWS ***,the soil moisture,snow water equivalent,and canopy water storage from the Global Land data Assimilation system were used to calculate the natural component of the TWS variations.A comparison of these results revealed the drivers of the changes in water *** results showed that:(1) overall TWS retrieved by GRACE decreased substantially as a consequence of human activity,while TWS obtained by WBF fluctuated periodically around zero under the impact of three natural drivers;(2) TWS in the NCP fell at a remarkable rate of-12.39 mm/y from 2004 to 2019,and the natural part of TWS experienced two significant declines,during 2004–2010 and 2013–2016;(3) variations in runoff,precipitation,and evapotranspiration from 2004 to 2019 were not significant,but human activities contributed more to the decreasing TWS than natural *** study provides a reference for water resource management and groundwater exploitation across the NCP under climate change.

关键词： North China Plain water storage GRACE human activities climatic factor

来源：评论

学校读者我要写书评

暂无评论

Magnetic Activity and Parameters of 43 Flare Stars in the GWAC Archive

引用

Research in Astronomy and Astrophysics 2023年第1期23卷 171-183页

作者：李广伟吴潮周桂萍杨晨黎华丽陈洁辛立平王竞哈斯铁尔哈尔肯马超红蔡洪波韩旭辉黄垒卢晓猛白建迎张旭康郝新丽王玉祥戴子高梁恩维孟小峰魏建彦 Key laboratory of Space Astronomy and Technology National Astronomical ObservatoriesChinese Academy of SciencesBeijing 100101China Key laboratory of Solar Activity National Astronomical ObservatoriesChinese Academy of SciencesBeijing 100101China Lab of Web and Mobile Data Management Renmin University of ChinaBeijing 100872China Guangxi Key Laboratory for Relativistic Astrophysics School of Physical Science and TechnologyGuangxi UniversityNanning 530004China School of Artificial Intelligence Beijing Normal UniversityBeijing100875China School of Astronomy and Space Science Nanjing UniversityNanjing 210093China Key Laboratory of Modern Astronomy and Astrophysics(Nanjing University) Ministry of EducationNanjing 210093China

In the archive of the Ground Wide Angle Camera(GWAC),we found 43 white light flares from 43 stars,among which,three are sympathetic or homologous flares,and one of them also has a quasi-periodic pulsation with a period of 13.0±1.5 *** these 43 flare stars,there are 19 new active stars and 41 stars that have available TESS and/or K2 light curves,from which we found 931 stellar *** also obtained rotational or orbital periods of 34 GWAC flare stars,of which 33 are less than 5.4 days,and ephemerides of three eclipsing binaries from these light *** with low resolution spectra from LAMOST and the Xinglong 2.16 m telescope,we found that L_(Hα)/L_(bol) are in the saturation region in the rotation-activity *** the LAMOST medium-resolution spectrum,we found that Star#3(HAT 178-02667)has double Hαemissions which imply it is a binary,and two components are both active *** stars have flare frequency distributions(FFDs)from TESS and/or K2 light *** FFDs show that the flares detected by GWAC can occur at a frequency of 0.5to 9.5 yr^(-1).The impact of flares on habitable planets was also studied based on these FFDs,and flares from some GWAC flare stars may produce enough energetic flares to destroy ozone layers,but none can trigger prebiotic chemistry on their habitable planets.

关键词： (stars:)binaries eclipsing-stars flare-stars low-mass-stars rotation

来源：评论

学校读者我要写书评

暂无评论

An Extended-Isomap for high-dimensional data accuracy and efficiency: a comprehensive survey

引用

Multimedia Tools and Applications 2024年第38期83卷 85523-85574页

作者： Yousaf, Mahwish Shakoor Khan, Muhammad Saadat Ullah, Shamsher Institute of Intelligence Machine Hefei Institutes of Physical Science Chinese Academy of Sciences Anhui Hefei230027 China Key Laboratory of Materials Physics Institute of Solid-State Physics Chinese Academy of Sciences Anhui Hefei230027 China National Engineering Laboratory for Big Data System Computing Technology Shenzhen University Guangdong Shenzhen518060 China

Manifold learning is a widely adopted nonlinear dimensionality reduction technique employed to discover low-dimensional representations from high-dimensional data and to explore the intrinsic data structure. It encompasses a range of nonlinear techniques, including Isomap, Local Linear Embedding (LLE), Local Tangent space Alignment (LTSA), and Multidimensional Scaling (MDS), among others. Isomap, as a popular method in manifold learning, has exhibited promising results. However, it encounters several challenges, such as the Topological Stability Problem, Shortest Path Distance, and noise sensitivity, which compromise its accuracy and elevate its computational demands. This survey paper presents three novel existing approaches to address these challenges and enhance the performance of Isomap. Firstly, the FastIsomap method, an existing approach that combines state-of-the-art algorithms like the KD tree and NN-descent to increase graph accuracy while reducing computational cost. Secondly, the FastIsomapVis method, also an existing technique, employs a hierarchical divide-and-conquer strategy employing the KD tree and Dijkstra Buckets Double, which enhances accuracy and reduces computational time in Isomap. Lastly, the NR-Isomap method, another existing approach, used the Local Tangent space Alignment algorithm to remove noise sensitivity and short-circuit edge problems. The existing methods improve the accuracy of Isomap but also streamline its computational requirements. Furthermore, this survey paper discusses the open issues and future research directions for Isomap and manifold learning methods. It can be a valuable resource for researchers seeking insights across diverse domains. © The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2024.

关键词： Clustering algorithms

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：