检索结果-内蒙古大学图书馆

Knowledge-Aware Multimodal Pre-Training for Fake News Detection

学校读者我要写书评

暂无评论

SSRN

SSRN 2024年

作者： Zhang, Litian Zhang, Xiaoming Zhou, Ziyi Zhang, Xi Li, Chaozhuo Yu, Philip S. School of Cyber Science and Technology Beihang University Beijing100191 China State Key Laboratory of Software Development Environment Beihang University Beijing100191 China Beijing University of Posts and Telecommunications Beijing China Department of Computer Science University of Illinois at Chicago Chicago United States

Amidst the rapid propagation of multimodal fake news across various social media platforms, the identification and filtering of disinformation have emerged as critical areas of academic research. A salient characteristic of fake news lies in its diversity, encompassing text-image incongruity, content-knowledge inconsistency, and content fabrication. However, existing endeavors are generally tailored to a specific subset of fake news, leading to limited universality. Moreover, these models primarily rely on scarce and exorbitant manually labeled annotations, which is incapable of providing sufficient learning signals to detect a variety of fake news. To address these challenges, we propose a novel knowledge-aware multimodal pre-training paradigm for fake news detection, dubbed KAMP. Our motivation lies in incorporating unsupervised correlations through pre-training tasks as complementary to alleviate the dependency on annotations. KAMP consists of a novel multimodal learning model and various delicate pre-training tasks to simultaneously capture valuable knowledge from single modality, multiple modalities, and background knowledge graphs. Our proposal undergoes comprehensive evaluation across two widely utilized datasets, and experimental results demonstrate the superiority of our proposal. © 2024, The Authors. All rights reserved.

关键词： Fake detection

Almost Optimal Time Lower Bound for Approximating Parameterized Clique, CSP, and More, under ETH

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Guruswami, Venkatesan Lin, Bingkai Ren, Xuandi Sun, Yican Wu, Kewen Simons Institute for the Theory of Computing Departments of EECS and Mathematics UC Berkeley United States State Key Laboratory for Novel Software Technology Nanjing University China Department of EECS UC Berkeley United States School of Computer Science Peking University China

The Parameterized Inapproximability Hypothesis (PIH), which is an analog of the PCP theorem in parameterized complexity, asserts that, there is a constant Ε > 0 such that for any computable function f : N → N, no f(k) · nO(1)-time algorithm can, on input a k-variable CSP instance with domain size n, find an assignment satisfying 1 − Ε fraction of the constraints. A recent work by Guruswami, Lin, Ren, Sun, and Wu (STOC’24) established PIH under the Exponential Time Hypothesis (ETH). In this work, we improve the quantitative aspects of PIH and prove (under ETH) that approximating sparse parameterized CSPs within a constant factor requires nk1−o(1) time. This immediately implies that, assuming ETH, finding a (k/2)-clique in an n-vertex graph with a k-clique requires nk1−o(1) time. We also prove almost optimal time lower bounds for approximating k-ExactCover and Max k-Coverage. Our proof follows the blueprint of the previous work to identify a "vector-structured" ETH-hard CSP whose satisfiability can be checked via an appropriate form of "parallel" PCP. Using further ideas in the reduction, we guarantee additional structures for constraints in the CSP. We then leverage this to design a parallel PCP of almost linear size based on Reed-Muller codes and derandomized low degree testing. Copyright © 2024, The Authors. All rights reserved.

关键词： Blueprints

Efficient Batch Verification and Privacy-Preserving Data Aggregation Scheme in V2G Networks

学校读者我要写书评

暂无评论

IEEE Transactions on Vehicular technology 2025年

作者： Wang, Congcong Wang, Chen Shen, Jian Vasilakos, Athanasios V. Wang, Bin Wang, Weizheng Department of Software Engineering Nanjing University of Information Science and Technology Nanjing210044 China the Zhejiang Key Laboratory of Digital Fashion and Data Governance the Zhejiang Provincial International Cooperation Base for Science and Technology on Cloud Computing Security and Data Aggregation Zhejiang Sci-Tech University Hangzhou310018 China Zhejiang Provincial Innovation Center of Advanced Textile Technology Shaoxing312000 China College of Computer Science and Information Technology IAU Saudi Arabia Grimstad4879 Norway Network and Data Security Hangzhou310053 China Computer Science Department City University of Hong Kong Hong Kong the Division of Computer Science University of Aizu Fukushima965–8580 Japan

In recent years, against the backdrop of smart grids, vehicle-to-grid (V2G) interaction has become a hot topic. With the increasing number of new energy vehicles, handling massive amounts of electricity transaction data has become an urgent problem in need of resolution. Ensuring the security of the vast quantities of electricity transaction data has emerged as a significant challenge. Therefore, two data aggregation schemes are proposed in this paper, both supporting batch verification and privacy protection in centralized and distributed scenarios, respectively. Firstly, an improved Paillier cryptosystem is utilized to achieve lightweight encryption in the proposed schemes. In addition, a novel distributed data aggregation structure that combines a large number of geographically distributed fog nodes with centralized cloud servers is introduced. This structure not only meets the requirements of lightweight encryption and resistance against single point failures but also reduces the communication complexity of distributed data aggregation to O(n). According to the security proofs and experimental analysis, it can be demonstrated that both proposed schemes are effective and secure, meeting the lightweight requirements of modern smart grid systems. © 2025 IEEE. All rights reserved,

关键词： Data aggregation

Hybrid Reinforced Medical Report Generation with M-Linear Attention and Repetition Penalty

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Xu, Wenting Xu, Zhenghua Chen, Junyang Qi, Chang Lukasiewicz, Thomas State Key Laboratory of Reliability and Intelligence of Electrical Equipment Tianjin Key Laboratory of Bioelectromagnetic Technology and Intelligent Health Hebei University of Technology China College of Computer Science and Software Engineering Shenzhen University China Department of Computer Science University of Oxford United Kingdom

To reduce doctors’ workload, deep-learning-based automatic medical report generation has recently attracted more and more research efforts, where deep convolutional neural networks (CNNs) are employed to encode the input images, and recurrent neural networks (RNNs) are used to decode the visual features into medical reports automatically. However, these state-of-the-art methods mainly suffer from three shortcomings: (i) incomprehensive optimization, (ii) low-order and unidimensional attention mechanisms, and (iii) repeated generation. In this article, we propose a hybrid reinforced medical report generation method with m-linear attention and repetition penalty mechanism (HReMRG-MR) to overcome these problems. Specifically, a hybrid reward with different weights is employed to remedy the limitations of single-metric-based rewards. We also propose a search algorithm with linear complexity to approximate the best weight combination. Furthermore, we use m-linear attention modules to explore high-order feature interactions and to achieve multi-modal reasoning, while a repetition penalty applies penalties to repeated terms during the model’s training process. Extensive experimental studies on two public datasets show that HReMRG-MR greatly outperforms the state-of-the-art baselines in terms of all metrics. We also conducted a series of ablation experiments to prove the effectiveness of all our proposed components. We also performed a reward search toy experiment to give evidence that our proposed search approach can significantly reduce the search time while approximating the best performance. Copyright © 2022, The Authors. All rights reserved.

关键词： Reinforcement learning

Effective, Platform-Independent GUI Testing via Image Embedding and Reinforcement Learning

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Yu, Shengcheng Fang, Chunrong Li, Xin Ling, Yuchen Chen, Zhenyu Su, Zhendong State Key Laboratory for Novel Software Technology Nanjing University Nanjing210093 China Department of Computer Science ETH Zurich Zurich8092 Switzerland

software applications (apps) have been playing an increasingly important role in various aspects of society. In particular, mobile apps and web apps are the most prevalent among all applications and are widely used in various industries as well as in people’s daily lives. To help ensure mobile and web app quality, many approaches have been introduced to improve app GUI testing via automated exploration, including random testing, model-based testing, learning-based testing, etc. Despite the extensive effort, existing approaches are still limited in reaching high code coverage, constructing high-quality models, and being generally applicable. Reinforcement learning-based approaches, as a group of representative and advanced approaches for automated GUI exploration testing, are faced with difficult challenges, including effective app state abstraction, reward function design, etc. Moreover, they heavily depend on the specific execution platforms (i.e., Android or Web), thus leading to poor generalizability and being unable to adapt to different platforms. This work specifically tackles these challenges based on the high-level observation that apps from distinct platforms share commonalities in GUI design. Indeed, we propose PIRLTest, an effective platform-independent approach for app testing. Specifically, PIRLTest utilizes computer vision and reinforcement learning techniques in a novel, synergistic manner for automated testing. It extracts the GUI widgets from GUI pages and characterizes the corresponding GUI layouts, embedding the GUI pages as states. The app GUI state combines the macroscopic perspective (app GUI layout) and the microscopic perspective (app GUI widget), and attaches the critical semantic information from GUI images. This enables PIRLTest to be platform-independent and makes the testing approach generally applicable on different platforms. PIRLTest explores apps with the guidance of a curiosity-driven strategy, which uses a Q-network to estimate the

关键词： Reinforcement learning

A Bivariate Control Strategy on Inductive Power Transfer Converter for Multi-Stage Constant Current Charging 48

学校读者我要写书评

暂无评论

A Bivariate Control Strategy on Inductive Power Transfer Con...

48th Annual Conference of the IEEE Industrial Electronics Society, IECON 2022

作者： Ding, Zhaoyi Iam, Io-Wa Ieong, Chi-Fong Lam, Chi-Seng State Key Laboratory of Analog and Mixed-Signal VLSI University of Macau Macau China Institute of Microelectronics University of Macau Macau China Department of Electrical and Computer Engineering Faculty of Science and Technology University of Macau Macau China

ISBN: (数字)9781665480253

ISBN: (纸本)9781665480253

Compared with the conventional constant current-constant voltage (CC-CV) charging profile of Li-ion battery, multi-stage constant current (MSCC) charging profile that composed of different constant current stages can improve the charging rate. This paper proposes a MSCC charging profile for an inductive power transfer (IPT) converter with LCC-S compensation topology. However, the wide load range variation during charging process impacts the transfer efficiency of the IPT converter. Thus, aiming to improve the transfer efficiency throughout the whole MSCC charging process, a bivariate operation approach is proposed with a variable inductance to achieve different constant current stages, while the voltage-controlled semi-active rectifier (VCSAR) is controlled to ensure the converter's efficiency optimization. Finally, simulation study is conducted based on PSIM software simulation to verify the validity and effectiveness of the bivariate control strategy and MSCC charging for the LCC-S compensated IPT converter. © 2022 IEEE.

关键词： Efficiency

Efficient Edge Intelligence under Clustering for UAV Swarm Networks

学校读者我要写书评

暂无评论

Efficient Edge Intelligence under Clustering for UAV Swarm N...

2021 IEEE International Conference on Space-Air-Ground Computing, SAGC 2021

作者： Qu, Yuben Dong, Chao Wu, Tao Zhuang, Yan Dai, Haipeng Wu, Fan Nanjing University of Aeronautics and Astronautics The Key Laboratory of Dynamic Cognitive System of Electromagnetic Spectrum Space Ministry of Industry and Information Technology China National University of Defense Technology China Shanghai Jiao Tong University Department of Computer Science and Engineering China Nanjing University State Key Laboratory for Novel Software Technology China

ISBN: (纸本)9781665400114

Unmanned aerial vehicles (UAVs) are expected to be widely used in many critical applications including monitoring, surveillance, urban target tracking, and delivery of goods, from military to civil fields. To realize that, empowering UAV swarm with edge intelligence is becoming appealing and inevitable. Federated learning (FL) as an emerging distributed collaborative machine learning (ML) without letting raw data out, could be suitable for UAV swarm networks but remains largely unexplored. In this paper, we first investigate the problem of efficient Edge Intelligence under Clustering for UAV swarm networks (e-EIC), i.e., how to jointly optimize the computation resource allocation and UAV clustering to minimize the total energy consumption of the UAV swarm for FL, subject to the training accuracy and latency constraints. We decompose the challenging e-EIC problem into two subproblems: computation resource allocation and UAV clustering, and accordingly propose an iterative algorithm based on the optimal policy of the first subproblem and local-search techniques for the second subproblem. Extensive performance evaluations demonstrate that our proposed algorithm can reduce more energy consumption of UAVs than three baseline algorithms, while satisfying the training accuracy and latency constraints. © 2021 IEEE.

关键词： Energy utilization

Demystifying What Code Summarization Models Learned

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Wang, Yu Wang, Ke Nanjing University China Visa Research United States State Key Laboratory for Novel Software Technology Department of Computer Science and Technology Nanjing University Jiangsu Nanjing210023 China Security Cryptography Visa Research Palo AltoCA United States

Study patterns that models have learned has long been a focus of pattern recognition research. Explaining what patterns are discovered from training data, and how patterns are generalized to unseen data are instrumental to understanding and advancing the pattern recognition methods. Unfortunately, the vast majority of the application domains deal with continuous data (i.e. statistical in nature) out of which extracted patterns can not be formally defined. For example, in image classification, there does not exist a principle definition for a label of cat or dog. Even in natural language, the meaning of a word can vary with the context it is surrounded by. Unlike the aforementioned data format, programs are a unique data structure with a well-defined syntax and semantics, which creates a golden opportunity to formalize what models have learned from source code. This paper presents the first formal definition of patterns discovered by code summarization models (i.e. models that predict the name of a method given its body), and gives a sound algorithm to infer a context-free grammar (CFG) that formally describes the learned patterns. We realize our approach in PATIC which produces CFGs for summarizing the patterns discovered by code summarization models. In particular, we pick two prominent instances, code2vec and code2seq, to evaluate PATIC. PATIC shows that the patterns extracted by each model are heavily restricted to local, and syntactic code structures with little to none semantic implication. Based on these findings, we present two example uses of the formal definition of patterns: a new method for evaluating the robustness and a new technique for improving the accuracy of code summarization models. Our work opens up this exciting, new direction of studying what models have learned from source code. © 2023, CC BY.

关键词： Deep neural networks

Learning Self-Target Knowledge for Few-Shot Segmentation

学校读者我要写书评

暂无评论

SSRN

SSRN 2023年

作者： Chen, Yadang Chen, Sihan Yang, Zhi-Xin Wu, Enhua Engineering Research Center of Digital Forensics Ministry of Education Nanjing University of Information Science and Technology Nanjing210044 China School of Computer Science Nanjing University of Information Science and Technology Nanjing210044 China State Key Laboratory of Internet of Things for Smart City Department of Electromechanical Engineering University of Macau 999078 China State Key Laboratory of Computer Science Institute of Software University of Chinese Academy of Sciences Beijing100190 China

Few-shot semantic segmentation uses a few annotated data of a specific class in the support set to segment the target of the same class in the query set. Most existing approaches fail to perform well when there are significant intra-class variances. This paper alleviates the problem by concentrating on mining the query image and using the support set as supplementary information. First, it proposes a Query Prototype Generation Module to generate a query foreground prototype from the query features. Specifically, we use both prototype-level and pixel-level similarity matching to generate two complementary initial prototypes, which we then integrate to create a discriminative query foreground prototype. Second, we propose a Support Auxiliary Refinement Module to further guide the final precise prediction of the query image by leveraging the target category information of the support set through step-by-step mining. Specifically, we generate a query-support mixture prototype based on the support prototype representation obtained using the attention mechanism. Then we generate a support supplement prototype to complement the missing information by encoding over the foreground regions that the query-support mixture prototype fails to segment out. Extensive experiments on PASCAL-5i and COCO-20i demonstrate that our model outperforms the prior works of few-shot segmentation. © 2023, The Authors. All rights reserved.

关键词： Semantics