检索结果-内蒙古大学图书馆

Cross-Modal Causal Representation Learning for Radiology Report Generation

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society 2025年 PP卷 PP页

作者： Weixing Chen Yang Liu Ce Wang Jiarui Zhu Guanbin Li Cheng-Lin Liu Liang Lin School of Computer Science and Engineering Sun Yat-sen University Guangzhou China Guangdong Key Laboratory of Big Data Analysis and Processing Guangzhou China School of Science Sun Yat-sen University Shenzhen China Department of Health Technology and Informatics The Hong Kong Polytechnic University Hung Hom Hong Kong State Key Laboratory of Multimodal Artificial Intelligence Systems Institute of Automation Chinese Academy of Sciences Beijing China School of Artificial Intelligence University of Chinese Academy of Sciences Beijing China

Radiology Report Generation (RRG) is essential for computer-aided diagnosis and medication guidance, which can relieve the heavy burden of radiologists by automatically generating the corresponding radiology reports according to the given radiology image. However, generating accurate lesion descriptions remains challenging due to spurious correlations from visual-linguistic biases and inherent limitations of radiological imaging, such as low resolution and noise interference. To address these issues, we propose a two-stage framework named Cross-Modal Causal Representation Learning (CMCRL), consisting of the Radiological Cross-modal Alignment and Reconstruction Enhanced (RadCARE) pre-training and the Visual-Linguistic Causal Intervention (VLCI) fine-tuning. In the pre-training stage, RadCARE introduces a degradation-aware masked image restoration strategy tailored for radiological images, which reconstructs high-resolution patches from low-resolution inputs to mitigate noise and detail loss. Combined with a multiway architecture and four adaptive training strategies (e.g., text postfix generation with degraded images and text prefixes), RadCARE establishes robust cross-modal correlations even with incomplete data. In the VLCI phase, we deploy causal front-door intervention through two modules: the Visual Deconfounding Module (VDM) disentangles local-global features without fine-grained annotations, while the Linguistic Deconfounding Module (LDM) eliminates context bias without external terminology databases. Experiments on IU-Xray and MIMIC-CXR show that our CMCRL pipeline significantly outperforms state-of-the-art methods, with ablation studies confirming the necessity of both stages. Code and models are available at.

关键词： Visualization Linguistics Radiology Heart Training Correlation Feature extraction Lesions Accuracy Lungs

来源：评论

学校读者我要写书评

暂无评论

Clifford circuit based heuristic optimization of fermion-to-qubit mappings

arXiv

引用

arXiv 2025年

作者： Yu, Jeffery Liu, Yuan Sugiura, Sho Van Voorhis, Troy Zeytinŏglu, Sina Joint Center for Quantum Information and Computer Science NIST University of Maryland College ParkMD20742 United States Joint Quantum Institute NIST University of Maryland College ParkMD20742 United States Department of Electrical and Computer Engineering North Carolina State University RaleighNC27606 United States Department of Computer Science North Carolina State University RaleighNC27606 United States Department of Physics North Carolina State University RaleighNC27606 United States Blocq Inc. 717 Market Street San FranciscoCA94103 United States Department of Chemistry Massachusetts Institute of Technology CambridgeMA02139 United States Institute for Theoretical Physics TU Wien Wiedner Hauptstraße 8-10/136 ViennaA-1040 Austria Physics and Informatics Laboratory NTT Research Inc. SunnyvaleCA94085 United States

Simulation of interacting fermionic Hamiltonians is one of the most promising applications of quantum computers. However, the feasibility of analysing fermionic systems with a quantum computer hinges on the efficiency of fermion-to-qubit mappings that encode non-local fermionic degrees of freedom in local qubit degrees of freedom. While recent works have highlighted the importance of designing fermion-to-qubit mappings that are tailored to specific problem Hamiltonians, the methods proposed so far are either restricted to a narrow class of mappings or they use computationally expensive and unscalable brute-force search algorithms. Here, we address this challenge by designing a heuristic numerical optimization framework for fermion-to-qubit mappings. To this end, we first translate the fermion-to-qubit mapping problem to a Clifford circuit optimization problem, and then use simulated annealing to optimize the average Pauli weight of the problem Hamiltonian. For all fermionic Hamiltonians we have considered, the numerically optimized mappings outperform their conventional counterparts, including ternary-tree-based mappings that are known to be optimal for single creation and annihilation operators. We find that our optimized mappings yield between 15% to 40% improvements on the average Pauli weight when the simulation Hamiltonian has an intermediate level of complexity. Most remarkably, the optimized mappings improve the average Pauli weight for 6×6 nearest-neighbor hopping and Hubbard models by more than 40% and 20%, respectively. Surprisingly, we also find specific interaction Hamiltonians for which the optimized mapping outperform any ternary-tree-based mapping. Our results establish heuristic numerical optimization as an effective method for obtaining mappings tailored for specific fermionic Hamiltonian. © 2025, CC BY.

关键词： Hamiltonians

来源：评论

学校读者我要写书评

暂无评论

How functional programming mattered

引用

National Science Review 2015年第3期2卷 349-370页

作者： Zhenjiang Hu John Hughes Meng Wang National Institute of Informatics Department of Computer Science and Engineering Chalmers University of Technology School of Computing University of Kent Department of Informatics SOKENDAI(Graduate University for Advanced Studies)

In 1989 when functional programming was still considered a niche topic, Hughes wrote a visionary paper arguing convincingly ‘why functional programming maters’. More than two decades have passed. Has functional programming really matered? Our answer is a resounding ‘Yes!’. Functional programming is now at the forefront of a new generation of programming technologies, and enjoying increasing popularity and inluence. In this paper, we review the impact of functional programming, focusing on how it has changed the way we may construct programs, the way we may verify programs, and fundamentally the way we may think about programs.

关键词： functional programming functional languages equational reasoning monad high order function

来源：评论

学校读者我要写书评

暂无评论

Methodology for optimizing electrical grounding grids in stratified soils using advanced calculation techniques and evolutionary algorithms

引用

Swarm and Evolutionary Computation 2025年 95卷

作者： Silva, Carlos Leandro Borges da Pires, Thyago Gumeratto Silva Filho, Antonio Marcelino Bulhões, Junio Santos Belo, Orlando M. Oliveira Rodrigues, Clóves Gonçalves Coimbra, Antonio Paulo Calixto, Wesley Pacheco Electrical Mechanical & Computer Engineering School Federal University of Goias Goiania Goias PC 74.605-220 Brazil Institute of Systems and Robotics Coimbra University CoimbraPC 3030-290 Portugal Department of Informatics School of Engineering University of Minho BragaPC 4710-057 Portugal Electrical Power Systems Study Group Federal Institute of Tocantins Palmas TocantinsPC 77.021-090 Brazil Graduate Program in Production and Systems Engineering Pontifical Catholic University of Goias Goiania Goias PC 74605-010 Brazil Mato Grosso Brazil Technology Research and Development Center Federal Institute of Goias Goiania Goias PC 74.055-110 Brazil

This paper presents a practical methodology for optimizing the geometry of electrical grounding grids at industrial frequencies of 50Hz and 60Hz, integrating advanced calculation techniques and evolutionary algorithms to improve the safety and operational performance of electrical grounding systems. The proposed approach is particularly beneficial for industrial automation and control systems, where effective grounding is necessary to maintain system reliability and prevent downtime. This methodology employs mathematical modeling and computational tools to optimize grid parameters, ensuring compliance with safety standards while reducing operational costs, thus contributing to the overall efficiency of automated systems in industrial environments. The study reports a reduction of up to 66% in the number of vertical rods and 40% in horizontal conductors compared to traditional methods. These results indicate that the proposed methodology can significantly reduce material usage and costs while maintaining electrical safety in accordance with regulatory standards, making it applicable to a wide range of industrial settings, including substations and automated facilities. © 2025 The Authors

关键词： Electric grounding

来源：评论

学校读者我要写书评

暂无评论

Cross-Layer Distributed Attack Detection Model for the IoT

引用

International Journal of Ambient Computing and Intelligence 2022年第1期13卷 1-17页

作者： Ahmed, Hassan I. Nasr, Abdurrahman A. Abdel-Mageid, Salah M. Aslan, Heba K. Informatics Department Electronics Research Institute Cairo Egypt Computer Engineering Faculty of Engineering Al-Azhar University Cairo Egypt Computer Engineering Department College of Computer Science and Engineering Taibah University Saudi Arabia Center of Informatics Science Faculty of Information Technology and Computer Science Nile University Egypt

The security of IoT that is based on layered approaches has shortcomings such as the redundancy, inflexibility, and inefficiently of security solutions. There are many harmful attacks in IoT networks such as DoS and DDoS attacks, which can compromise the IoT architecture in all layers. Consequently, cross layer approach is proposed as an effective and practical security defending mechanism. Cross-layer distributed attack detection model (CLDAD) is proposed to enhance security solutions for IoT environments. CLDAD presents a general detection method of DDoS in sensing layer, network layer, and application layer. CLDAD is based on big data analytics techniques, which enable the detection process to be performed in a distributed way, so the model can detect DDoS attacks in any layer on-the-fly and the model support the scalability of the IoT environment. CLDAD is tested based on three datasets, namely artificial jamming attack dataset, BoT-IoT dataset, and BoT-IoT-based HTTP. The results showed that the proposed model is efficient in detecting attacks in the three layers of the IoT. Copyright © 2022, IGI Global.

关键词： Big data

来源：评论

学校读者我要写书评

暂无评论

A Machine Learning Approach for Analyzing Texture and Morphometry Markers of Glioblastoma TERT Promoter Mutations in DAPI Fluorescence Images

A Machine Learning Approach for Analyzing Texture and Morpho...

引用

Medical Imaging 2025: Digital and Computational Pathology

作者： Afsari, Fatemeh Naglah, Ahmed Gupta, Akshita Lohaan, Harshit Onubogu, Ugoma Janiszewska, Michalina Sarder, Pinaki Division of Nephrology Hypertension and Renal Transplantation-Quantitative Health Section Department of Medicine College of Medicine University of Florida GainesvilleFL32603 United States Department of Health Outcomes and Biomedical Informatics College of Medicine University of Florida GainesvilleFL32611 United States Department of Computer and Information Science and Engineering University of Florida GainesvilleFL32611 United States Department of Molecular Medicine The Herbert Wertheim UF Scripps Institute for Biomedical Innovation and Technology JupiterFL33458 United States

ISBN: (纸本)9781510686045

Advanced molecular and imaging techniques, such as STAR-FISH, are frequently required to identify mutations in tissue specimens to properly diagnose Glioblastoma (GBM), an aggressive type of brain tumor. Mutations to the telomerase reverse transcriptase promoter (TERTp) are common in cancer and present in approximately 80% of all GBM tumors. We hypothesize that machine learning can stratify cells by their TERTp mutation status using DAPI fluorescence images since the STAR-FISH technique can be costly and inaccessible for some healthcare facilities. In this study, we present a machine learning-based approach to recognize nuclei harboring TERTp mutations solely based on the DAPI fluorescence images of GBM tissues. The proposed pipeline integrates a convolutional neural network (CNN)-based deep learning model to segment cells, extract hand-engineered features, and feed them into fully connected (FC) layers within the neural network or other machine learning classifiers to identify TERTp mutant cells. This approach involves extracting DAPI features, including both morphological and textural image features, followed by an ensemble of machine learning modules to predict TERTp mutant cells. We evaluate our approach on a cohort of 18 patients (70 pairs of DAPI and STAR-FISH processed images). Our experimental results show that the FC layers within the neural network achieve the highest AUC, F1-Score, and recall of 0.77, 0.69, and 0.86, respectively, with the integrated CNN model achieving a Dice score of 0.83, accurately segmenting cells. Our work is one of the earliest attempts that we are aware of to use machine-learning models for TERTp mutation identification using DAPI fluorescence images. Our results indicate the possibility of improved GBM diagnosis accuracy. To improve patient diagnosis and treatment planning for GBM, future research aims will expand the cohort and enhance the models to provide more precise and diagnostically valuable models for the interpretation o

关键词： Image segmentation

来源：评论

学校读者我要写书评

暂无评论

Efficient redundant assignments under Fault-Tolerance constraints 3rd

引用

3rd International Workshop on Randomization and Approximation Techniques in computer Science and 2nd International Workshop on Approximation Algorithms for Combinatorial Optimization Problems, RANDOM-APPROX 1999

作者： Fotakis, Dimitris A. Spirakis, Paul G. Computer Engineering and Informatics Department University of Patras Rion Patras265 00 Greece Computer Technology Institute CTI Kolokotroni 3 Patras262 21 Greece

ISBN: (纸本)3540663290

We consider the problem of computing minimum congestion, fault-tolerant, redundant assignments of messages to faulty parallel delivery channels. In particular, we are given a set M of faulty channels, each having an integer capacity ci and failing independently with probability fi. We are also given a set of messages to be delivered over M, and a fault-tolerance constraint (1—∈), and we seek a redundant assignment _ that minimizes congestion Cong(), i.e. the maximum channel load, subject to the constraint that, with probability no less than (1—∈), all the messages have a copy on at least one active channel. We present a 4-approximation algorithm for identical capacity channels and arbitrary message sizes, and a 2 (Formula is presented)-approximation algorithm for related capacity channels and unit size messages. Both algorithms are based on computing a collection of disjoint channel subsets such that, with probability no less than (1—∈), at least one channel is active in each subset. The objective is to maximize the sum of the minimum subset capacities. Since the exact version of this problem is NP-complete, we present a 2-approximation algorithm for identical capacities, and a (8 + o(1))-approximation algorithm for arbitrary capacities. © Springer-Verlag Berlin Heidelberg 1999.

关键词： Fault tolerance

来源：评论

学校读者我要写书评

暂无评论

Improved detection mechanism in MANETs using adaptive IP tracing mechanism

引用

International Journal of Intelligent Networks 2020年 1卷 102-106页

作者： Mamade, Bayisa Kune Dhas, C. Suresh Gnana School of Informatics and Electrical Engineering Institute of Technology Ambo University Ethiopia Department of Computer Science Institute of Technology Ambo University Ethiopia

The network has enabled enormous volumes of services without any restrictions, different users need to have access to the service provided are more focused by malicious users. It is imperative to identify malicious entry or access to services, hence this paper discusses the various aspects of malicious access due to DDoS attacks and on how to detect it also this method uses the decisions made by hackers to manage the Traceback information science. In this paper, we consider the flash crowds based on the difference between DDoS attacks, a flash crowd is used to mitigate the DDoS attack. There is a tendency to see information on flux entropy variations on routers’ short-term and store them. Once the victim has been aware of a DDoS attack, the victim starts pushback tracing. The initial Traceback rule identifies its upstream routers where attacks have taken place and transmits the requests for Traceback to the upstream connected routers. If the flow rate is low, the low detection rule is initiated to trace the attackers and therefore the study is evaluated based on the instances where the packets are attacked by malicious users. The evaluation is conducted based on whether the packet traversed is malicious or normal and finally the accuracy of IP Traceback techniques and time complexity are tested. © 2020 The Author(s)

关键词： Entropy

来源：评论

学校读者我要写书评

暂无评论

Making P-space smart: Integrating IoT technologies in a multi-office environment

引用

5th International Conference on Mobile Wireless Middleware, Operating Systems, and Applications, Mobilware 2012

作者： Akribopoulos, Orestis Amaxilatis, Dimitrios Georgitzikis, Vasileios Logaras, Marios Keramidas, Vasileios Kontodimas, Konstantinos Lagoudianakis, Evangelos Nikoloutsakos, Nikolaos Papoutsakis, Vasileios Prevezanos, Ioannis Pyrgeris, Georgios Tsampas, Stylianos Voutsas, Vasileios Chatzigiannakis, Ioannis Computer Technology Institute and Press Computer Engineering and Informatics Department University of Patras Greece

ISBN: (纸本)9783642366598

Internet of Things technologies are considered the next big step in Smart Building installations. Although such technologies have been widely studied in simulation and experimental scenarios it is not so obvious how problems of real world installations should be dealt with. In this work we deploy IoT devices for sensing and control in a multi-office space and employ technologies such as CoAP, RESTful interfaces and Semantic Descriptions to integrate them with the Web. We report our research goals, the challenges we faced, the decisions we made and the experience gained from the design, deployment and operation of all the hardware and software components that compose our system. © 2013 ICST institute for computer Science, Social informatics and Telecommunications engineering.

关键词： Intelligent buildings

来源：评论

学校读者我要写书评

暂无评论

A mobile web map application for route discovery with device independent characteristics 07

A mobile web map application for route discovery with device...

引用

2007 Euro American Conference on Telematics and Information Systems, EATIS 2007

作者： Andriakopoulou, Eirini Garofalakis, John Plessas, Athanasios University of Patras Computer Engineering and Informatics Department Research Academic Computer Technology Institute

ISBN: (纸本)9781595935984

As mobile devices are becoming increasingly powerful and ubiquitous, new possibilities to get route guidance are created, allowing users to save valuable time and effort. However, mobile devices still face severe constraints that limit the possibilities that a mobile route finding application may offer. In this paper, we present a device independent mobile application that enables mobile users to receive route directions to their handheld device. This application is an extension to an existing mobile map delivery system, the m-CHARTIS system. We focus on the main features and we analyse the device independent character of the application. Implementation details are presented and finally directions for future work are proposed.

关键词： Mobile computing

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：