检索结果-内蒙古大学图书馆

Multi-Scale Feature Guided Transformer for Image Inpainting

IET Image Processing 2025年第1期19卷

作者： Huang, Zeji Lu, Huanda Yu, Xin Xiao, Hui School of Computer Science and Technology Zhejiang Sci-Tech University Hangzhou China School of Computer and Data Engineering NingboTech University Ningbo China

In recent years, image restoration has witnessed remarkable advancements. However, reconstructing visually plausible textures while preserving global structural coherence remains a persistent challenge. Existing convolutional neural network (CNN)-based approaches are inherently limited by their local receptive fields, often struggling to capture global structure. Previously proposed methods mostly focus on structural priors to address the limitation of CNN's receptive field, but we believe that texture priors are also critical factors that influence the quality of image inpainting. To tackle semantic inconsistency and texture blurriness in current methods, we introduce a novel multi-stage restoration framework. Specifically, our architecture incorporates a dual-stream U-Net with attention mechanisms to extract multi-scale features. The mixed attention-gated feature fusion module exchanges and combines structure and texture features to generate multi-scale fused feature maps, which are progressively merged into the decoder to guide the Transformer to generate more realistic images. Additionally, we propose a feature selection feedforward network to replace traditional MLPs in Transformer blocks for adaptive feature refinement. Extensive experiments on CelebA-HQ and Paris StreetView datasets demonstrate superior performance both qualitatively and quantitatively compared to state-of-the-art methods. © 2025 The Author(s). IET Image Processing published by John Wiley & Sons Ltd on behalf of The Institution of Engineering and Technology.

关键词： Multilayer neural networks

来源：评论

学校读者我要写书评

暂无评论

Improved Safety and 3D Scanning with Human-Robot Collaboration 25

Improved Safety and 3D Scanning with Human-Robot Collaborati...

引用

25th International Conference on Digital Image Computing: Techniques and Applications, DICTA 2024

作者： Wright, Madison Nam, Karlym Tong, Jinguang Muthu, Sundaram Andersson, Lars Nguyen, Chuong Csiro Imaging & Computer Vision - Data61 Canberra Australia College of Eng. Computing & Cybernetics The Australian National University Australia School of Electrical Eng. & Computer Science The University of Queensland Australia

ISBN: (纸本)9798350379037

In many technical working environments, humans need robots to perform, or assist with, repetitive tasks. Since robots remain a tool from our perspective, the question is how to safely use this tool to its full potential. We focus on two use cases: one is safety and the other is knowledge sharing. First, collaborative robots, or cobots, have been used more and more in factory work cells due to their ability to stop and prevent serious injuries when they collide to human operators. It would be even better if the collision is avoided in the first place with robotic vision. Second, robots could carry out tasks but may lack the knowledge or skill to deal with difficult situations, for example: completing a full 3D scanning of an object. In this paper, we present two approaches to improve safety and 3D scanning in a research lab environment. The setup is a work cell where a human operator works with a robot arm which is equipped with an RGBD camera and a gripper to manipulate and scan objects of interest. A second RGBD camera looks at the work cell for human gestures and their body movement. Robotic Operating System (ROS) version 2 is employed for hardware-software integration. One of our approaches allows the robot to track and follow objects of interest as they are moved around on a table, and stop when it detects a human hand in the way. Our other approach allows the robot to recognise human gestures to capture RGBD data from multiple views for 3D scanning and flip the object for scanning the hidden side. The captured data is then combined to have a more complete 3D information of the object. We demonstrate the effectiveness of our robot performing these tasks and the improvement in work safety and the quality of 3D object reconstruction. © 2024 IEEE.

关键词： Gesture recognition

来源：评论

学校读者我要写书评

暂无评论

Stitching, Fine-tuning, Re-training: A SAM-enabled Framework for Semi-supervised 3D Medical Image Segmentation

引用

IEEE Transactions on Medical Imaging 2025年 PP卷 PP页

作者： Li, Shumeng Qi, Lei Yu, Qian Huo, Jing Shi, Yinghuan Gao, Yang Nanjing University State Key Laboratory for Novel Software Technology China Nanjing University National Institute of Healthcare Data Science China School of Computer Science and Engineering China Shandong Women's University School of Data and Computer Science China

Segment Anything Model (SAM) fine-tuning has shown remarkable performance in medical image segmentation in a fully supervised manner, but requires precise annotations. To reduce the annotation cost and maintain satisfactory performance, in this work, we leverage the capabilities of SAM for establishing semi-supervised medical image segmentation models. Rethinking the requirements of effectiveness, efficiency, and compatibility, we propose a three-stage framework, i.e., Stitching, Fine-tuning, and Re-training (SFR). The current fine-tuning approaches mostly involve 2D slice-wise fine-tuning that disregards the contextual information between adjacent slices. Our stitching strategy mitigates the mismatch between natural and 3D medical images. The stitched images are then used for fine-tuning SAM, providing robust initialization of pseudo-labels. Afterwards, we train a 3D semi-supervised segmentation model while maintaining the same parameter size as the conventional segmenter such as V-Net. Our SFR framework is plug-and-play, and easily compatible with various popular semi-supervised methods. We also develop an extended framework SFR+ with selective fine-tuning and re-training through confidence estimation. Extensive experiments validate that our SFR and SFR+ achieve significant improvements in both moderate annotation and scarce annotation across five datasets. In particular, SFR framework improves the Dice score of Mean Teacher from 29.68% to 74.40% with only one labeled data of LA dataset. The code is available at https://***/ShumengLI/SFR. © 1982-2012 IEEE.

关键词： Image segmentation

来源：评论

学校读者我要写书评

暂无评论

Provably Secure Conditional-Privacy Access Control Protocol for Intelligent Customers-Centric Communication in VANET

引用

IEEE Transactions on Consumer Electronics 2024年第1期70卷 1747-1756页

作者： Saleem, Muhammad Asad Li, Xiong Mahmood, Khalid Shamshad, Salman Ayub, Muhammad Faizan Bashir, Ali Kashif Omar, Marwan University of Electronic Science and Technology of China School of Computer Science and Engineering Sichuan Chengdu12599 China National Yunlin University of Science and Technology Graduate School of Intelligent Data Science Douliu64002 Taiwan The University of Lahore Department of Software Engineering Lahore54590 Pakistan Manchester Metropolitan University Department of Computing and Mathematics ManchesterM15 6BH United Kingdom Woxsen University Woxsen School of Business Hyderabad502345 India Lebanese American University Department of Computer Science and Mathematics Beirut1102 2801 Lebanon Illinois Institute of Technology Information Technology and Management ChicagoIL60555 United States

Globally, the development of Intelligent Cyber-Physical Transportation Systems (ICTS) aims to tackle several challenges, including reducing traffic accidents and fuel usage, alleviating congestion, shortening travel time, and enhancing overall transportation safety. These systems leverage advanced customer-centric communication and networked control methods, such as inter-vehicle, vehicle-to-roadside (V2R), and vehicle-to-vehicle (V2V) communication through the use of vehicular ad hoc networks (VANETs) to cover all aspects of transportation-based information. In existing systems, once devices are registered with a Trusted Authority (TA), subsequent authentication still relies on the TA's assistance. However, these devices typically remain stationary, and frequent interaction with the TA becomes impractical and costly in highly mobile VANET environments. To address this challenge, we propose a secure access control protocol with conditional privacy for VANETs. Unlike other protocols, our protocol does not require the TA's involvement during authentication between vehicles and RSUs. Additionally, our protocol leverages pseudonym mechanisms to provide conditional privacy, enabling legitimate vehicles to remain anonymous while malicious ones can be tracked. Our proposed scheme is supported by both formal and informal security analyses and has been shown to be secure against several known attacks in VANETs. Furthermore, compared to relevant studies, our protocol achieves 30.2505% of efficiency in terms of computation cost and 11.09276% of efficiency in terms of communication cost, respectively. © 1975-2011 IEEE.

关键词： Vehicular ad hoc networks

来源：评论

学校读者我要写书评

暂无评论

Order Doesn’t Matter, But Reasoning Does: Training LLMs with Order-Centric Augmentation

arXiv

引用

arXiv 2025年

作者： He, Qianxi He, Qianyu Liang, Jiaqing Xiao, Yanghua Zhou, Weikang Sun, Zeye Yu, Fei Shanghai Key Laboratory of Data Science School of Computer Science Fudan University China School of Data Science Fudan University China Ant Group China

Logical reasoning is essential for large language models (LLMs) to ensure accurate and coherent inference. However, LLMs struggle with reasoning order variations and fail to generalize across logically equivalent transformations. LLMs often rely on fixed sequential patterns rather than true logical understanding. To address this issue, we introduce an order-centric data augmentation framework based on commutativity in logical reasoning. We first randomly shuffle independent premises to introduce condition order augmentation. For reasoning steps, we construct a directed acyclic graph (DAG) to model dependencies between steps, which allows us to identify valid reorderings of steps while preserving logical correctness. By leveraging order-centric augmentations, models can develop a more flexible and generalized reasoning process. Finally, we conduct extensive experiments across multiple logical reasoning benchmarks, demonstrating that our method significantly enhances LLMs’ reasoning performance and adaptability to diverse logical structures. We release our codes and augmented data in https://***/r/Order-Centric-data-Augmentation-822C/. Copyright © 2025, The Authors. All rights reserved.

关键词： Directed graphs

来源：评论

学校读者我要写书评

暂无评论

Injective coloring of edge corona product of graphs 8

Injective coloring of edge corona product of graphs

引用

2024 8th International Conference on Mechanics, Mathematics and Applied Physics, ICMMAP 2024

作者： Zhao, Wenying Tian, Shuangliang School of Mathematics and Computer Science Northwest for Minzu University Lanzhou730030 China Key Laboratory of Streaming Data Computing Technologies and Applications Northwest for Minzu University Lanzhou730030 China

The injective chromatic number χi(G) of a graph G is the minimum number of colors needed to color the vertices of G such that different vertices with the same neighbor are colored differently. In this here, we get some exact values of the injective chromatic number of the edge corona product of some graphs. © Published under licence by IOP Publishing Ltd.

关键词： Graph theory

来源：评论

学校读者我要写书评

暂无评论

Applying Hybrid Clustering in Pulsar Candidate Sifting with Multi-modality for FAST Survey

引用

Research in Astronomy and Astrophysics 2024年第3期24卷 283-296页

作者： Zi-Yi You Yun-Rong Pan Zhi Ma Li Zhang Shuo Xiao Dan-Dan Zhang Shi-Jun Dang Ru-Shuang Zhao Pei Wang Ai-Jun Dong Jia-Tao Jiang Ji-Bing Leng Wei-An Li Si-Yao Li School of Physics and Electronic Science Guizhou Normal UniversityGuiyang 550025China Guizhou Provincial Key Laboratory of Radio Astronomy and Data Processing Guizhou Normal UniversityGuiyang 550025China College of Big Data and Information Engineering Guizhou UniversityGuiyang 550025China CAS Key Laboratory of FAST National Astronomical ObservatoriesChinese Academy of SciencesBeijing 100101China Key Laboratory of Information and Computing Guizhou Province Guizhou Normal UniversityGuiyang 550001China School of Big data and Computer Science Guizhou Normal UniversityGuiyang 550025China Guizhou Software Engineering Research Center Guiyang 550000China

Pulsar search is always the basis of pulsar navigation,gravitational wave detection and other research ***,the volume of pulsar candidates collected by the Five-hundred-meter Aperture Spherical radio Telescope(FAST)shows an explosive growth rate that has brought challenges for its pulsar candidate filtering ***,the multi-view heterogeneous data and class imbalance between true pulsars and non-pulsar candidates have negative effects on traditional single-modal supervised classification *** this study,a multi-modal and semi-supervised learning based on a pulsar candidate sifting algorithm is presented,which adopts a hybrid ensemble clustering scheme of density-based and partition-based methods combined with a feature-level fusion strategy for input data and a data partition strategy for *** on both High Time Resolution Universe SurveyⅡ(HTRU2)and actual FAST observation data demonstrate that the proposed algorithm could excellently identify pulsars:On HTRU2,the precision and recall rates of its parallel mode reach0.981 and 0.988 *** FAST data,those of its parallel mode reach 0.891 and 0.961,meanwhile,the running time also significantly decreases with the increment of parallel nodes within ***,we can conclude that our algorithm could be a feasible idea for large scale pulsar candidate sifting for FAST drift scan observation.

关键词： methods data analysis-surveys-methods numerical

来源：评论

学校读者我要写书评

暂无评论

Zero Watermarking Algorithm for Medical Image Based on Resnet50-DCT

引用

computers, Materials & Continua 2023年第4期75卷 293-309页

作者： Mingshuai Sheng Jingbing Li Uzair Aslam Bhatti Jing Liu Mengxing Huang Yen-Wei Chen School of Information and Communication Engineering Hainan UniversityHaikou570228China State Key Laboratory of Marine Resource Utilization in the South China Sea Hainan UniversityHaikou570228China School of Computer Science and Technology Hainan UniversityHaikou570228China Research Center for Healthcare Data Science Zhejiang LabHangzhou311121China Graduate School of Information Science and Engineering Ritsumeikan UniversityKyoto5258577Japan

Medical images are used as a diagnostic tool, so protecting theirconfidentiality has long been a topic of study. From this, we propose aResnet50-DCT-based zero watermarking algorithm for use with medicalimages. To begin, we use Resnet50, a pre-training network, to draw out thedeep features of medical images. Then the deep features are transformedby DCT transform and the perceptual hash function is used to generatethe feature vector. The original watermark is chaotic scrambled to get theencrypted watermark, and the watermark information is embedded into theoriginal medical image by XOR operation, and the logical key vector isobtained and saved at the same time. Similarly, the same feature extractionmethod is used to extract the deep features of the medical image to be testedand generate the feature vector. Later, the XOR operation is carried outbetween the feature vector and the logical key vector, and the encryptedwatermark is extracted and decrypted to get the restored watermark;thenormalized correlation coefficient (NC) of the original watermark and therestored watermark is calculated to determine the ownership and watermarkinformation of the medical image to be tested. After calculation, most ofthe NC values are greater than 0.50. The experimental results demonstratethe algorithm’s robustness, invisibility, and security, as well as its ability toaccurately extract watermark information. The algorithm also shows goodresistance to conventional attacks and geometric attacks.

关键词： Medical images deep residual network resnet50-DCT privacy protection robustness security

来源：评论

学校读者我要写书评

暂无评论

Order Matters: Investigate the Position Bias in Multi-constraint Instruction Following

arXiv

引用

arXiv 2025年

作者： Zeng, Jie He, Qianyu Ren, Qingyu Liang, Jiaqing Xiao, Yanghua Zhou, Weikang Sun, Zeye Yu, Fei Shanghai Key Laboratory of Data Science School of Computer Science Fudan University China School of Data Science Fudan University China Ant Group China

Real-world instructions with multiple constraints pose a significant challenge to existing large language models (LLMs). An observation is that the LLMs exhibit dramatic performance fluctuation when disturbing the order of the incorporated constraints. Yet, none of the existing works has systematically investigated this position bias problem in the field of multi-constraint instruction following. To bridge this gap, we design a probing task where we quantitatively measure the difficulty distribution of the constraints by a novel Difficulty Distribution Index (CDDI). Through the experimental results, we find that LLMs are more performant when presented with the constraints in a "hard-to-easy" order. This preference can be generalized to LLMs with different architecture or different sizes of parameters. Additionally, we conduct an explanation study, providing an intuitive insight into the correlation between the LLM’s attention and constraint orders. Our code and dataset are publicly available at https://***/meowpass/ PBIF. Copyright © 2025, The Authors. All rights reserved.

关键词：

来源：评论

学校读者我要写书评

暂无评论

OblivTime: Oblivious and Efficient Interval Skyline Query Processing Over Encrypted Time-Series data

引用

IEEE Transactions on Services Computing 2025年第3期18卷 1602-1617页

作者： Ouyang, Huajie Zheng, Yifeng Wang, Songlei Hua, Zhongyun Harbin Institute of Technology School of Computer Science and Technology Guangdong Shenzhen518055 China The Hong Kong Polytechnic University Department of Electrical and Electronic Engineering Hong Kong Shenzhen University National Engineering Laboratory for Big Data System Computing Technology Shenzhen518055 China

Time-series data is prevalent in many applications like smart homes, smart grids, and healthcare. And it is now increasingly common to store and query time-series data in the cloud. Despite the benefits, data privacy concerns in such outsourced services are pressing, making it imperative to embed privacy assurance mechanisms from the outset. Most existing related works have been focused on querying for different types of aggregate statistics. In this paper, we instead focus on the secure support for advanced interval skyline queries, which allow to identify time series that are not dominated by any other time series within a query time interval. This is valuable for time-series data analytics in applications like remote health monitoring (e.g., identifying patients with high heart rates in a certain week). We present OblivTime, a new system framework for oblivious and efficient interval skyline query processing over encrypted time-series data. OblivTime is built from a synergy of time-series data analytics, lightweight cryptography, and GPU parallel computing, achieving stronger security guarantees and lower online query latency over the state-of-the-art prior work. Extensive experiments demonstrate that OblivTime can achieve up to 666\times speedup in online query latency over the state-of-the-art prior work. © 2008-2012 IEEE.

关键词： Structured Query Language

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：