检索结果-内蒙古大学图书馆

17th Asian Conference on Computer Vision, ACCV 2024

作者： Anh, Duy Le Dinh Tran, Kim Hoang Nguyen, Quang-Thuc Le, Ngan Hoang University of Arkansas FayettevilleAR United States Faculty of Information Technology and Software Engineering Lab University of Science VNU-HCM Ho Chi Minh City Viet Nam Vietnam National University Ho Chi Minh City Viet Nam

ISBN: (纸本)9789819609161

Despite recent progress, Multi-Object Tracking (MOT) continues to face significant challenges, particularly its dependence on prior knowledge and predefined categories, complicating the tracking of unfamiliar objects. Generic Multiple Object Tracking (GMOT) emerges as a promising solution, requiring less prior information. Nevertheless, existing GMOT methods, primarily designed as OneShot-GMOT, rely heavily on initial bounding boxes and often struggle with variations in viewpoint, lighting, occlusion, and scale. To overcome the limitations inherent in both MOT and GMOT when it comes to tracking objects with specific generic attributes, we introduce Grounded-GMOT, an innovative tracking paradigm that enables users to track multiple generic objects in videos through natural language descriptors. Our contributions begin with the introduction of the G2MOT dataset, which includes a collection of videos featuring a wide variety of generic objects, each accompanied by detailed textual descriptions of their attributes. Following this, we propose a novel tracking method, KAM-SORT, which not only effectively integrates visual appearance with motion cues but also enhances the Kalman filter. KAM-SORT proves particularly advantageous when dealing with objects of high visual similarity from the same generic category in GMOT scenarios. Through comprehensive experiments, we demonstrate that Grounded-GMOT outperforms existing OneShot-GMOT approaches. Additionally, our extensive comparisons between various trackers highlight KAM-SORT’s efficacy in GMOT, further establishing its significance in the field. Project page: https://***/G2MOT. The source code and dataset will be made publicly available. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Kalman filters

来源：评论

学校读者我要写书评

暂无评论

Separate Source Channel Coding Is Still What You Need:An LLM-Based Rethinking

引用

ZTE Communications 2025年第1期23卷 30-44页

作者： REN Tianqi LI Rongpeng ZHAO Mingmin CHEN Xianfu LIU Guangyi YANG Yang ZHAO Zhifeng ZHANG Honggang College of Information Science and Electronic Engineering Zhejiang UniversityHangzhou 310027China Shenzhen CyberAray Network Technology Co. Ltd.Shenzhen 518000China China Mobile Research Institute Beijing 100053China The Internet of Things Thrust The Hong Kong University of Science and Technology(Guangzhou)Guangzhou 511453China Zhejiang Lab Hangzhou 311121China Faculty of Data Science City University of MacaoMacao 999078China

Along with the proliferating research interest in semantic communication(Sem Com),joint source channel coding(JSCC)has dominated the attention due to the widely assumed existence in efficiently delivering information ***,this paper challenges the conventional JSCC paradigm and advocates for adopting separate source channel coding(SSCC)to enjoy a more underlying degree of freedom for *** demonstrate that SSCC,after leveraging the strengths of the Large Language Model(LLM)for source coding and Error Correction Code Transformer(ECCT)complemented for channel coding,offers superior performance over *** proposed framework also effectively highlights the compatibility challenges between Sem Com approaches and digital communication systems,particularly concerning the resource costs associated with the transmission of high-precision floating point *** comprehensive evaluations,we establish that assisted by LLM-based compression and ECCT-enhanced error correction,SSCC remains a viable and effective solution for modern communication *** other words,separate source channel coding is still what we need.

关键词： separate source channel coding(SSCC) joint source channel coding(JSCC) end-to-end communication system Large Language Model(LLM) lossless text compression Error Correction Code Transformer(ECCT)

来源：评论

学校读者我要写书评

暂无评论

Conformal Graph-level Out-of-distribution Detection with Adaptive Data Augmentation 25

Conformal Graph-level Out-of-distribution Detection with Ada...

引用

34th ACM Web Conference, WWW 2025

作者： Lin, Xixun Cao, Yanan Sun, Nan Zou, Lixin Zhou, Chuan Zhang, Peng Zhang, Shuai Zhang, Ge Wu, Jia Institute of Information Engineering Chinese Academy of Sciences Beijing China Institute of Information Engineering Chinese Academy of Sciences School of Cyber Security University of Chinese Academy of Sciences Beijing China School of Cyber Science and Engineering Wuhan University Wuhan China Academy of Mathematics and Systems Science CAS School of Cyber Security University of Chinese Academy of Sciences Beijing China Cyberspace Institute of Advanced Technology Guangzhou University Guangzhou China Academy of Mathematics and Systems Science CAS Beijing China School of Computer Science and Technology Donghua University Shanghai China Department of Computing Macquarie University Sydney Australia

ISBN: (纸本)9798400712746

Graph-level out-of-distribution (OOD) detection, which attempts to identify OOD graphs originated from an unknown distribution, is a vital building block for safety-critical applications in Web and society. Current approaches concentrate on how to learn better graph representations, but fail to provide any statistically guarantee on detection results, therefore impeding their deployments in the scenario where detection errors would result in serious consequences. To overcome this critical issue, we propose the Conformal Graph-level Out-of-distribution Detection (CGOD), extending the theory of conformal prediction to graph-level OOD detection with a rigorous control over the false positive rate. In CGOD, we develop a new aggregated non-conformity score function based on the proposed adaptive data augmentation. Through the guidance from two designed metrics, i.e., score consistency and representation diversity, our augmentation strategy can generate multiple non-conformity scores, and aggregating these generated non-conformity scores together is robust to the misleading information. Meanwhile, our score function can perceive the subsequent process of conformal inference, enabling the aggregated non-conformity score to be adaptive to different input graphs and deriving a more accurate score estimation. We conduct experiments on multiple real-world datasets with different empirical settings. Extensive results and model analyses demonstrate the superior performance of our approach over several competitive baselines. © 2025 Copyright held by the owner/author(s).

关键词： Network theory (graphs)

来源：评论

学校读者我要写书评

暂无评论

Anomaly Detection with LWE Encrypted Control

arXiv

引用

arXiv 2025年

作者： Alisic, Rijad Kim, Junsoo Sandberg, Henrik the Division of Decision and Control Systems KTH Royal Institute of Technology Sweden the Department of Electrical and Information Engineering Seoul National University of Science and Technology Korea Republic of

Detecting attacks using encrypted signals is challenging since encryption hides its information content. We present a novel mechanism for anomaly detection over Learning with Errors (LWE) encrypted signals without using decryption, secure channels, nor complex communication schemes. Instead, the detector exploits the homomorphic property of LWE encryption to perform hypothesis tests on transformations of the encrypted samples. The specific transformations are determined by solutions to a hard lattice-based minimization problem. While the test’s sensitivity deteriorates with suboptimal solutions, similar to the exponential deterioration of the (related) test that breaks the cryptosystem, we show that the deterioration is polynomial for our test. This rate gap can be exploited to pick parameters that lead to somewhat weaker encryption but large gains in detection capability. Finally, we conclude the paper by presenting a numerical example that simulates anomaly detection, demonstrating the effectiveness of our method in identifying attacks. Copyright © 2025, The Authors. All rights reserved.

关键词： Cryptography

来源：评论

学校读者我要写书评

暂无评论

As an AI Language Model, "Yes I Would Recommend Calling the Police": Norm Inconsistency in LLM Decision-Making 24

As an AI Language Model, "Yes I Would Recommend Calling the ...

引用

Proceedings of the 2024 AAAI/ACM Conference on AI, Ethics, and Society

作者： Shomik Jain D. Calacci Ashia Wilson Institute for Data Systems and Society Massachusetts Institute of Technology College of Information Sciences and Technology Penn State University Department of Electrical Engineering and Computer Science Massachusetts Institute of Technology

We investigate the phenomenon of norm inconsistency: where LLMs apply different norms in similar situations. Specifically, we focus on the high-risk application of deciding whether to call the police in Amazon Ring home surveillance videos. We evaluate the decisions of three state-of-the-art LLMs - GPT-4, Gemini 1.0, and Claude 3 Sonnet - in relation to the activities portrayed in the videos, the subjects' skin-tone and gender, and the characteristics of the neighborhoods where the videos were recorded. Our analysis reveals significant norm inconsistencies: (1) a discordance between the recommendation to call the police and the actual presence of criminal activity, and (2) biases influenced by the racial demographics of the neighborhoods. These results highlight the arbitrariness of model decisions in the surveillance context and the limitations of current bias detection and mitigation strategies in normative decision-making.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Multi-Layer Secured Edge-AI Enabled Model for Consumer Health systems

引用

IEEE Transactions on Consumer Electronics 2025年

作者： Rehman, Amjad Haseeb, Khalid Saba, Tanzila Jeon, Gwanggil Al-Otaibi, Shaha Lab Riyadh12435 Saudi Arabia Incheon National University College of Information Technology Department of Embedded Systems Engineering Incheon22012 Korea Republic of Princess Nourah bint Abdulrahman University College of Computer and Information Sciences Department of Information Systems Riyadh11671 Saudi Arabia

The Internet of Medical Things (IoMT) and edge computing technologies have significantly advanced the healthcare industry by enabling real-time monitoring and data processing at the network edges. Medical devices are combined with consumer electronics to measure the physiological factors of the human body and transfer them to cloud storage for further analysis and processing. However, optimizing routing to account for limited processing power and network constraints of the consumer healthcare system poses a significant research challenge. Moreover, due to the restricted environment, several privacy and security issues arise with integrating heterogeneous technologies over wireless systems. In this work, we proposed a multi-layer secured-edged-enabled AI model for health systems to provide real-time data analysis and transmission decisions at the edges. Firstly, the intelligent algorithm employs dynamic and trusted routing strategies at the edges to control the data flow and minimize the network congestion on consumers’ data. Secondly, in the event of data and integrity breaches, randomized security patterns are investigated to build trust among devices using edge management and raise the degree of data privacy. The randomized security patterns improve the analysis of data security by integrating the concept of rotating encryption keys and their management. Such a paradigm decreases the probabilities of network threats and generates unpredictable patterns for increasing the resilience of IoT systems in the existence of harmful devices. The proposed model is tested using various simulation based experiments, revealing its significant improvement over existing studies for packet drop ratio, data delay, system uptime, data breaches, and security overhead. © 1975-2011 IEEE.

关键词： Electronic health record

来源：评论

学校读者我要写书评

暂无评论

SVD-Guided Multimodal Feature Fusion for Emotion Recognition from Facial Videos

引用

IEEE Transactions on Affective Computing 2025年

作者： Bao, Jindi Qian, Jianjun Yang, Jian Nanjing University of Science and Technology Pca Lab Key Lab of Intelligent Percept. and Syst. for High-Dimensional Information of Ministry of Education School of Computer Science and Engineering Nanjing210094 China

Multimodal emotion recognition based on facial videos aims to extract features from different modalities to identify human emotions. The previous work focus on designing various fusion schemes to combine heterogeneous modal data. However, most studies have overlooked the role of different modalities in emotion recognition and have not fully utilized the intrinsic connections between modalities. Furthermore, the multimodal data from facial videos also contain various distractions bad for emotion analysis. How to reduce the impact of distractions and enable a model to mine effective information for emotion recognition from different modalities is still a challenge problem. To address above issue, we propose a SVD-guided multimodal feature fusion method based on facial video for emotion recognition, which uses a hierarchical fusion mechanism and adopts different loss strategies at each level to learn multimodal feature representation. Specifically, we fuse the facial expression and rPPG signal (or Point-of-Gaze) by using the weak supervision strategy and contrastive learning. Subsequently, the fused feature of facial expression and rPPG signal and the fused feature of facial expression and Point-of-Gaze are combined together to construct the unified multimodal feature matrix. Based on this, Singular Value Decomposition (SVD) is used to refine the redundancy information caused by the multimodal fusion and guide the neural network to learn discriminative emotion feature. At the same time, a consistent loss is developed to enhance the multimodal representation. Experiments on three public datasets show that the proposed method achieves better results over the compared methods. © 2010-2012 IEEE.

关键词： Emotion Recognition

来源：评论

学校读者我要写书评

暂无评论

An Efficient and Scalable Byzantine Fault Tolerant Consensus for Vehicular Networks

引用

IEEE Transactions on Vehicular technology 2025年

作者： Deshmukh, Ameya Baiju, Dheeraj Atish, Atish Alladi, Tejasvi Yu, F. Richard BITS-Pilani Department of Computer Science and Information Systems Pilani333031 India BITS-Pilani Department of Electrical and Electronics Engineering Pilani333031 India Carleton University Department of Information Technology OttawaON Canada

Vehicular networks represent a new distributed system paradigm that requires robust fault tolerance to ensure reliable operation. As a burgeoning area of research, the scalability and optimization of consensus mechanisms for these networks are critical. Traditional Byzantine Fault Tolerant (BFT) algorithms like PBFT are not inherently optimized for the localized needs of vehicular networks, suffering from scalability issues due to their global nature and high messaging complexity. In response, we introduce a two-tiered consensus framework that refines PBFT for the specific context of vehicular networks. By organizing nodes into clusters based on geographic proximity, our approach reduces messaging complexity from O(n2) to O(n1.5), significantly improving scalability. The framework distinguishes between local and global state transitions, adding two phases to the PBFT protocol to manage these efficiently. This tailored consensus process aligns with the localized communication patterns of vehicular networks, enhancing both efficiency and scalability. The framework addresses the critical challenges of traditional BFT algorithms in vehicular networks, offering a solution that is both scalable and resilient. It is a step toward enabling vehicular networks to fulfil their potential as a reliable component of modern distributed systems. © 1967-2012 IEEE.

关键词： Blockchain Byzantine Fault Tolerance Distributed consensus Vehicular networks

来源：评论

学校读者我要写书评

暂无评论

EvoPath: Evolutionary Meta-path Discovery with Large Language Models for Complex Heterogeneous information Networks

arXiv

引用

arXiv 2025年

作者： Liu, Shixuan Cheng, Haoxiang Wang, Yunfei He, Yue Fan, Changjun Liu, Zhong Laboratory for Big Data and Decision College of Systems Engineering National University of Defense Technology Hunan Changsha China National Key Laboratory of Information Systems Engineering College of Systems Engineering National University of Defense Technology Hunan Changsha China Department of Computer Science and Technology Tsinghua University Beijing China

Heterogeneous information Networks (HINs) encapsulate diverse entity and relation types, with meta-paths providing essential meta-level semantics for knowledge reasoning, although their utility is constrained by discovery challenges. While Large Language Models (LLMs) offer new prospects for meta-path discovery due to their extensive knowledge encoding and efficiency, their adaptation faces challenges such as corpora bias, lexical discrepancies, and hallucination. This paper pioneers the mitigation of these challenges by presenting EvoPath, an innovative framework that leverages LLMs to efficiently identify high-quality meta-paths. EvoPath is carefully designed, with each component aimed at addressing issues that could lead to potential knowledge conflicts. With a minimal subset of HIN facts, EvoPath iteratively generates and evolves meta-paths by dynamically replaying meta-paths in the buffer with prioritization based on their scores. Comprehensive experiments on three large, complex HINs with hundreds of relations demonstrate that our framework, EvoPath, enables LLMs to generate high-quality meta-paths through effective prompting, confirming its superior performance in HIN reasoning tasks. Further ablation studies validate the effectiveness of each module within the framework. © 2025, CC BY-NC-ND.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

Enhancing surface detection: A comprehensive analysis of various YOLO models

引用

Heliyon 2025年第3期11卷 e42433页

作者： Deepti Raj, G. Prabadevi, B. School of Computer Science Engineering and Information Systems Vellore Institute of Technology University Tamilnadu Vellore 632014 India

Material defects can significantly affect strength, durability and overall quality. Complex backgrounds and variations in steel surface images often hinder productivity and quality in industrial environments. Accurate defect detection becomes difficult due to small target size and unclear features. However, implementing accurate and automated object detection algorithms mitigates these challenges, allowing errors or defects to be detected before processing. Version 5 of You Only Look Once (YOLO), a precisely optimized learning model, has undergone extensive testing on steel strip datasets, providing effective solutions for recognition and detection in industry environments. This study presents an improved YOLOv5 detection model, exploiting the efficient channel attention (ECA) and coordinated attention (CoordAtt) mechanisms. Our results show notable improvements, with the ECA hybrid attention mechanism achieving 2–4 times faster inference times while maintaining high accuracy. Additionally, CoordAtt integration minimizes the parameter count by 25 % and gives higher accuracy on one of the datasets. Comparative analysis with YOLOv6, YOLOv7, and YOLOv8 demonstrates the superior accuracy of the enhanced YOLOv5 model on the NEU-DET and GC10-DET steel strip benchmark datasets, highlighting its effectiveness in detecting and timely recognition of actual defects. © 2025 The Authors

关键词： Attention mechanisms Convolution neural networks Defect detection LeakyReLU Object detection YOLOv5

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：