检索结果-内蒙古大学图书馆

39th Annual AAAI Conference on Artificial Intelligence, AAAI 2025

作者： Feng, Zhida Chen, Li Sun, Yuenan Liu, Jiaxiang Feng, Shikun School of Computer Science and Technology Wuhan University of Science and Technology China Hubei Province Key Laboratory of Intelligent Information Processing and Real-time Industrial System Wuhan University of Science and Technology China Baidu Inc China

ISBN: (纸本)157735897X

ControlNet has significantly advanced controllable image generation by integrating dense conditions (such as depth and canny edges) with text-to-image diffusion models. However, ControlNet’s integration requires an additional amount nearly equal to half of the base diffusion model’s parameters, making it inefficient. To address this, we introduce Simple-ControlNet, an efficient and streamlined network for controllable text-to-image generation. It employs a single-scale projection layer to incorporate condition information into the denoising U-Net. It is supplemented by Low-Rank Adapter (LoRA) parameters to facilitate condition learning. Impressively, Simple-ControlNet requires fewer than 3 million parameters for the control mechanism, substantially less than the 300 million needed by ControlNet. Our extensive experiments confirm that Simple-ControlNet matches and surpasses ControlNet’s performance across a broad range of tasks and base diffusion models, showcasing its utility and efficiency. © 2025, Association for the Advancement of Artificial Intelligence (***). All rights reserved.

关键词： Controllability

来源：评论

学校读者我要写书评

暂无评论

Monocular Depth Estimation with Boundary Attention Mechanism and Shifted Window Adaptive Bins

SSRN

引用

SSRN 2024年

作者： Liang, Mengnan Wang, Congcong Zhao, Meng Shi, Fan Zhang, Chao Han, Yilin School of Computer Science and Engineering Tianjin University of Technology Tianjin300384 China Laboratory of Computer Vision System of Ministry of Education Tianjin University of Technology Tianjin300384 China

Monocular depth estimation is a classic research topic in computer vision. In recent years, development of Convolutional Neural Networks (CNNs) has facilitated significant breakthroughs in this field. However, there still are two challenges: 1) The network struggles to effectively fuse edge features in the feature fusion stage, which ultimately results in the loss of structure or distorted boundaries of objects in the scene. 2) Classification based studies typically depend on Transformers for global modeling, a process that often introduces substantial performance overhead. In this paper, we propose two modules to address the aforementioned issues. The first module is the Boundary Attention Module (BAM), which leverages the attention mechanism to enhance the ability of the network to perceive object boundaries during the feature fusion stage. In addition, to mitigate the performance overhead resulting from predicting adaptive bins, we propose a Shift Window Adaptive Bins (SWAB) module to reduce the amount of calculation in global modeling. The proposed method is evaluated on two public datasets, NYU Depth V2 and KITTI, and achieves state-of-the-art (SOTA) performance. © 2024, The Authors. All rights reserved.

关键词： computer vision

来源：评论

学校读者我要写书评

暂无评论

Prior Knowledge-Driven Hybrid Prompter Learning for RGB-Event Tracking

引用

IEEE Transactions on Circuits and systems for Video technology 2025年

作者： Wang, Mianzhao Shi, Fan Cheng, Xu Chen, Shengyong Tianjin University of Technology Engineering Research Center of Learning-Based Intelligent System Ministry of Education Key Laboratory of Computer Vision and System School of Computer Science and Engineering Tianjin300384 China Technical University of Denmark Department of Technology Management and Economics Kongens Lyngby Denmark

Event data can asynchronously capture variations in light intensity, thereby implicitly providing valuable complementary cues for RGB-Event tracking. Existing methods typically employ a direct interaction mechanism to fuse RGB and event data. However, due to differences in imaging mechanisms, the representational disparity between these two data types is not fixed, which can lead to tracking failures in certain challenging scenarios. To address this issue, we propose a novel prior knowledge-driven hybrid prompter learning framework for RGB-Event tracking. Specifically, we develop a frame-event hybrid prompter that leverages prior tracking knowledge from the foundation model as intermediate modal support to mitigate the heterogeneity between RGB and event data. By leveraging its rich prior tracking knowledge, the intermediate modal reduces the gap between the dense RGB and sparse event data interactions, effectively guiding complementary learning between modalities. Meanwhile, to mitigate the internal learning disparities between the lightweight hybrid prompter and the deep transformer model, we introduce a pseudo-prompt learning strategy that lies between full fine-tuning and partial fine-tuning. This strategy adopts a divide-and-conquer approach to assign different learning rates to modules with distinct functions, effectively reducing the dominant influence of RGB information in complex scenarios. Extensive experiments conducted on two public RGB-Event tracking datasets show that the proposed HPL outperforms state-of-the-art tracking methods, achieving exceptional performance. © 1991-2012 IEEE.

关键词： Object detection

来源：评论

学校读者我要写书评

暂无评论

Hyperbit: A Financial Temporal Knowledge Graph Data Storage system

SSRN

引用

SSRN 2022年

作者： Yuan, Pingpeng Han, Sheng Zang, Shaoqi Shi, Xuanhua Jin, Hai 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 Wuhan430074 China

In the field of finance, data has the characteristics of temporal multi-frequency and heterogeneous high dimension, and the traditional knowledge graph is not suitable to express the temporal relation of financial data, so the financial temporal knowledge graph has become a research hotspot recently. There are a large number of temporal edges in the financial temporal knowledge graph. The use of general graph database management system to store financial knowledge graph data has problems such as large storage scale and slow update speed, and it cannot realize the traceability of financial data, so it cannot meet the needs of the financial field. In order to solve these problems, HyperBit, a storage system for efficient storage and management of financial data, is designed and implemented. Firstly, a temporal quad storage structure is proposed. Secondly, the Log Structured Cumulative Update (LSCU) is proposed to update financial temporal data efficiently. The update log module is designed in the system. When there is data to update, it is not directly updated in the data storage of the system, but recorded in the update log, and the system updates the content of the update log cumulatively. The system manages update logs by partition and chunk, which makes the log storage compact and independent. Besides, it is convenient for parallel processing and it improves the data update speed. Thirdly, a Bidirectional Backtracking by Final and Optimal Root (BBFOR) method is proposed to meet the requirements of efficient traceability of financial data. Unlike traditional version tracing method, BBFOR adopt both the snapshot and incremental methods which balance the space and time cost of version switching, and BBROF consider the version of the access frequency characteristic and set up the final root and the optimal root. Besides, BBROF adopt bidirectional backtracking technology to improve the efficiency of the version switching. Experimental results show that the storage cost

关键词： Switching

来源：评论

学校读者我要写书评

暂无评论

Unknown-Aware Bilateral Dependency Optimization for Defending Against Model Inversion Attacks

引用

IEEE Transactions on Pattern Analysis and Machine Intelligence 2025年 PP卷 PP页

作者： Peng, Xiong Liu, Feng Wang, Nannan Lan, Long Liu, Tongliang Cheung, Yiu-Ming Han, Bo Hong Kong Baptist University Department of Computer Science Hong Kong The University of Melbourne School of Computing and Information Systems Australia Xidian University State Key Laboratory of Integrated Services Networks School of Telecommunications Engineering Shaanxi Xi'an710071 China National University of Defense Technology State Key Laboratory of High Performance Computing School of Computer Hunan Changsha410000 China The University of Sydney Sydney AI Centre School of Computer Science Faculty of Engineering Australia

By abusing access to a well-trained classifier, model inversion (MI) attacks pose a significant threat as they can recover the original training data, leading to privacy leakage. Previous studies mitigated MI attacks by imposing regularization to reduce the dependency between input features and outputs during classifier training, a strategy known as unilateral dependency optimization. However, this strategy contradicts the objective of minimizing the supervised classification loss, which inherently seeks to maximize the dependency between input features and outputs. Consequently, there is a trade-off between improving the model's robustness against MI attacks and maintaining its classification performance. To address this issue, we propose the bilateral dependency optimization strategy (BiDO), a dual-objective approach that minimizes the dependency between input features and latent representations, while simultaneously maximizing the dependency between latent representations and labels. BiDO is remarkable for its privacy-preserving capabilities. However, models trained with BiDO exhibit diminished capabilities in out-of-distribution (OOD) detection compared to models trained with standard classification supervision. Given the open-world nature of deep learning systems, this limitation could lead to significant security risks, as encountering OOD inputs—whose label spaces do not overlap with the in-distribution (ID) data used during training-is inevitable. To address this, we leverage readily available auxiliary OOD data to enhance the OOD detection performance of models trained with BiDO. This leads to the introduction of an upgraded framework, unknown-aware BiDO (BiDO+), which mitigates both privacy and security concerns. As a highlight, with comparable model utility, BiDO-HSIC+ reduces the FPR95 by 55.02% and enhances the AUCROC by 9.52% compared to BiDO-HSIC, while also providing superior MI robustness. © 1979-2012 IEEE.

关键词： Differential privacy

来源：评论

学校读者我要写书评

暂无评论

Digital Human Techniques for Education Reform 24

Digital Human Techniques for Education Reform

引用

7th International Conference on Educational technology Management, ICETM 2024

作者： Guo, Peirong Zhang, Qi Tian, Chunwei Xue, Wanli Feng, Xiaocheng School of Software Northwestern Polytechnical University Shaanxi Xi’an China School of Economics and Management Harbin Institute of Technology at Weihai Shandong Weihai China Shenzhen Research Institute of Northwestern Polytechnical University Northwestern Polytechnical University Guangdong Shenzhen China School of Computer Science and Engineering Tianjin University of Technology Tianjin China Research Center for Social Computing and Information Retrieval Harbin Institute of Technology Heilongjiang Harbin China Guangdong Key Laboratory of Intelligent Information Processing Shenzhen Key Laboratory of Media Security Shenzhen University Guangdong Shenzhen China

ISBN: (纸本)9798400717468

The rapid evolution of artificial intelligence, big data, and generative AI models has ushered in significant transformations across various sectors, including education. Digital Human Technique, an innovative technology grounded in advanced computer science and artificial intelligence, is reshaping educational paradigms by enabling virtual humans to simulate human behavior, express emotions, and interact with users. This paper explores the application of Digital Human Technique in education reform, focusing on creating immersive, intelligent classroom experiences that foster meaningful interactions between teachers and students. We define Digital Human Technique and delve into its key technical components such as character modeling and rendering, natural language processing, computer vision, and augmented reality technologies. Our methodology involves analyzing the role of educational digital humans created through these technologies, assessing their impact on educational processes, and examining various application scenarios in educational reform. Results indicate that Digital Human Technique significantly enhances the learning experience by enabling personalized teaching, increasing engagement, and fostering emotional connections. Educational digital humans serve as virtual teachers, interactive learning aids, and facilitators of emotional interaction, effectively addressing the challenges of traditional educational methods. They also promote a deeper understanding of complex concepts through simulated environments and interactive digital content. © 2024 Copyright held by the owner/author(s).

关键词： Digital elevation model

来源：评论

学校读者我要写书评

暂无评论

DeepRank: Test Case Prioritization for Deep Neural Networks 35

DeepRank: Test Case Prioritization for Deep Neural Networks

引用

35th International Conference on Software Engineering and Knowledge Engineering, SEKE 2023

作者： Li, Wei Zhang, Zhiyi Jian, Yifan Liu, Chen Huang, Zhiqiu Collage of Computer Science and Technology Nanjing University of Aeronautics and Astronautics Nanjing China Collaborative Innovation Center of Novel Software Technology and Industrialization Nanjing China Science and Technology on Reactor System Design Technology Laboratory Nuclear Power Institute of China Chengdu China School of Marxism Yangzhou University Yangzhou China Ministry Key Laboratory for Safety-Critical Software Development and Verification Nanjing University of Aeronautics and Astronautics Nanjing China

Deep neural networks (DNNs) have been widely used in safety-critical fields such as autonomous driving and medical diagnosis. However, DNNs are easily disturbed to make wrong decisions, which may lead to loss of life or property. Therefore, it is vital to test DNN adequately. In practice, to reveal the incorrect behavior of DNN and improve its robustness, testers usually need massive labeled data to test and optimize DNN. However, labeling test inputs to detect the correctness of DNN predictions is an expensive and time-consuming task that even affects the efficiency of DNN testing. To relieve the labeling-cost problem, we propose DeepRank, a test case prioritization technique based on cross-entropy loss. The key idea of DeepRank is that the higher the loss value of a test case relative to the DNN, the more likely it is to be mispredicted and the more conducive it is to improve the robustness of the DNN through retraining. Therefore, the cross-entropy loss value can be used for test case prioritization. We experimentally validate our approach on two datasets and three DNNs models. The experimental results demonstrate that DeepRank is significantly better than existing test case prioritization methods regarding fault-revealing capability and retraining effectiveness. © 2023 Knowledge systems Institute Graduate school. 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

CADEC: A Combinatorial Auction for Dynamic Distributed DNN Inference Scheduling in Edge-Cloud Networks

引用

IEEE Transactions on Mobile computing 2025年

作者： Xu, Xiaolong Hu, Yuhao Cui, Guangming Qi, Lianyong Dou, Wanchun Nanjing University of Information Science and Technology School of Software Nanjing210044 China Nanjing University of Information Science and Technology Jiangsu Province Engineering Research Center of Advanced Computing and Intelligent Services School of Software China Nanjing University State Key Laboratory for Novel Software Technology China College of Computer science and technology Qingdao China Georgia State University Department of Computer Science United States

Deep Neural Network (DNN) Inference, as a key enabler of intelligent applications, is often computation-intensive and latency-sensitive. Combining the advantages of cloud computing (abundant computing resources) and edge computing (fast transmission), edge-cloud collaborative DNN inference is a powerful solution to these problems. However, in edge-cloud networks with heterogeneous resources, how to obtain reasonable decisions on server selection, model partition and resource allocation for efficient distributed DNN inference is a hard challenge. Furthermore, it is non-trivial to design suitable resource prices to maximize the social welfare. These challenges even escalate in dynamic edge-cloud networks where decisions should be generated as soon as each user arrives without future information. Therefore, we design a combinatorial auction for dynamic distributed DNN inference scheduling, named CADEC. CADEC first constructs a bid set for each user based on convex optimization theory for optimal solution searching. Next, prices of resources in the edge-cloud network are adjusted according to changes in supply-demand relationship, and whether to admit the request of each user is decided. Finally, the dynamic distributed inference scheduling decisions are generated through the primal-dual algorithm to maximize the social welfare. Theoretical analysis shows the good competitive ratio and polynomial time complexity of CADEC. Results of simulation experiments present that CADEC improves social welfare by up to 46% compared with state-of-the-art distributed DNN inference schemes. © 2002-2012 IEEE.

关键词： Convex optimization

来源：评论

学校读者我要写书评

暂无评论

First wetland mapping at 10-m spatial resolution in South America using multi-source and multi-feature remote sensing data

引用

science China Earth sciences 2024年第10期67卷 3252-3269页

作者： Weiwei SUN Gang YANG Yuling HUANG Dehua MAO Ke HUANG Lin ZHU Xiangchao MENG Tian FENG Chao CHEN Yong GE Department of Geography and Spatial Information Techniques Ningbo UniversityNingbo 315211China State Key Laboratory of Resources and Environmental Information System Institute of Geographic Sciences&Natural Resources ResearchChinese Academy of SciencesBeijing 100101China Natural Resources and Planning Bureau Shanghai Pudong New DistrictShanghai 200100China Key Laboratory of Wetland Ecology and Environment Northeast Institute of Geography and AgroecologyChinese Academy of SciencesChangchun 130102China Faculty of Electrical Engineering and Computer Science Ningbo UniversityNingbo 315211China School of Geography Science and Geomatics Engineering Suzhou University of Science and TechnologySuzhou 215009China

Wetland degradation has been accelerating in recent years globally. Accurate information on the geographic distribution and categories of wetlands is essential for their conservation and management. Despite being the world′s fourth largest continent, South America has limited research on wetland mapping, and there is currently no available map that provides comprehensive information on wetland distribution and categories in the region. To address this issue, we used Sentinel-1, Sentinel-2 and SRTM data, developed a sample collection method and a wetland mapping method with a collection of multi-source features such as optical features, polarization features and shape features for South American wetlands. We produced a 10-m resolution wetland map based on the Google Earth Engine(GEE) platform. Our Level-1 wetland cover map accurately captured six wetland sub-categories with an overall accuracy of 96.24% and a kappa coefficient of 0.8649, while our Level-2 water cover map included five sub-categories with an overall accuracy of 97.23% and a kappa coefficient of0.9368. The results show that the total area of existing wetlands in South America is approximately 1,737,000 km~2, which is6.8% of the total land area. Among the ten wetland categories, shallow sea had the largest area(960,527.4 km~2), while aquaculture ponds had the smallest area 1513.6 km~2. Swamp had the second largest area(306,240.1 km~2). Brazil, Argentina,Venezuela, Bolivia, and Colombia were found to have the largest wetland areas, with Brazil and Colombia having the most diverse wetland categories. This product can serve as baseline data for subsequent monitoring, management, and conservation of South American wetlands.

关键词： Wetland mapping Google Earth Engine Sentinel imagery South America

来源：评论

学校读者我要写书评

暂无评论

Deep Learning Techniques in Mobile Edge computing for Internet of Medical Things 3

Deep Learning Techniques in Mobile Edge Computing for Intern...

引用

3rd International Conference on Emerging Smart Technologies and Applications, eSmarTA 2023

作者： Alawadhi, Abdulwadood Ahmad, R. Badlishah Almogahed, Abdullah Abrar, Ahmad Department of Computers and Information Technology Yemen Universiti Malaysia Perlis Faculty of Electronic Engineering and Technology Perlis Malaysia Universiti Tun Hussein Onn Malaysia Faculty of Computer Science and Information Technology Johor Parit Raja Malaysia School of Computing Universiti Utara Malaysia InterNetWorks Research Laboratory Sintok-Kedah Malaysia

ISBN: (纸本)9798350305333

The fast expansion of the Internet of Medical Things (IoMT) has resulted in a ubiquitous home health diagnostic network. High patient demand results in high costs, short latency, and communication overload. As a result, 6G is the next generation of IoT, IoMT, and cellular networks, intending to considerably improve the quality of smart healthcare services through high throughput and decreased latency. So far, adopting cloud computing for time-critical applications and decreasing access delays to resources is difficult. Deep learning has been extensively employed to extract characteristics from complicated networks as artificial intelligence technology has advanced. On the other hand, deep learning models are often run in cloud computing data centres with tremendous processing resources. Conventional cloud computing systems primarily rely on the network, which has significant latency and security and privacy issues. One of the successful solutions is the Mobile Edge computing (MEC) paradigm, which brings cloud computing services closer to the edge network and uses available resources. Mobile edge computing places computer and storage nodes near mobile devices at the Internet's edge, leading to considerable savings in system operating time, memory cost, and power usage. Deep learning is used in mobile edge computing to forecast changes in demand based on daily patient behaviours. It also prepares the network by scaling up network resources as required. This study aims to discuss IoMT and identfy the problems in deep learning for mobile edge computing technology and its applications. © 2023 IEEE.

关键词： Deep learning

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：