检索结果-内蒙古大学图书馆

MSMFT: Multi-Stream Multimodal Factorized Transformer for Human Activity Recognition

IEEE Sensors Journal 2025年第6期25卷 10402-10416页

作者： Zhou, Xianfa Yuan, Jiabin Fan, Lili Niu, Xuewei Zha, Keke Liu, Xu Nanjing University of Aeronautics and Astronautics Department of Computer Science and Technology Nanjing210016 China Jinling Institute of Technology Department of Computer Engineering Nanjing211169 China

Existing video-based human activity recognition (HAR) methods are susceptible to challenges such as lighting variations and occlusions in complex environments. Wearable sensors can effectively mitigate these issues. The fusion of the two can reduce environmental impacts, thus enhancing accuracy. However, the heterogeneity between the two data types presents a challenge for fusion. To address this, we propose an architecture called multi-stream multimodal factorized transformer (MSMFT), which leverages the aggregation capabilities of transformers to achieve alignment and fusion of multimodal features without relying on specific feature extractors. First, we designed a statistical feature projection (SFP) method to effectively tokenize wearable sensor data for transformer models. Second, we introduced temporal embedding (TE) to capture local temporal feature variations. Then, we incorporated a factorized transformer architecture that includes temporal and spatial transformer encoders (STEs) to extract global spatiotemporal features fully. Finally, to avoid the issue of modality entanglement caused by deep fusion, we constructed a multi-stream architecture comprising a video stream, wearable sensors stream, and fusion stream. This design allows for the extraction of both independent and interactive modality features. The performance of the MSMFT method was evaluated on three multimodal datasets: UP-Fall, UTD-MHAD, and MMAct. The results show that MSMFT performs better than the state-of-the-art, with a 16.83% improvement in cross-scene F1-score on the MMAct dataset. © 2001-2012 IEEE.

关键词： Wearable sensors

来源：评论

学校读者我要写书评

暂无评论

ProxaDyn: A Proximity-Aware Dynamic Caching Approach for Named Data Networks

引用

IEEE Transactions on Network science and engineering 2025年第3期12卷 2360-2372页

作者： Kumari, Matta Krishna Tripathi, Nikhil Joshi, Piyush Indian Institute of Information Technology Computer Science and Engineering Group Sri City517 646 India Department of Computer Science and Engineering Dhanbad826 004 India

Named Data Network (NDN), a future Internet architecture is introduced to address the shortcomings of the current Internet architecture. NDN supports in-network caching to facilitate scalable content distribution and enhance overall network performance. However, the known NDN caching strategies suffer from a few common drawbacks, such as inefficient cache utilization, high content redundancy, and overhead due to lookup repetition. To address these issues, in this paper, we propose a novel caching strategy called ProxaDyn for efficient content lookup, placement, and replacement. During the content lookup phase, ProxaDyn interacts exclusively with the router responsible for caching a particular content. This eliminates interaction with other intermediate routers, thereby significantly reducing content access latency. For content placement, ProxaDyn strategically selects an on-path router based on content popularity. Popular content is placed in the cache of a router closer to the consumer, while less popular content is cached in a router away from the consumer. This approach significantly improves the cache hits and reduces the access latency. We test ProxaDyn over a diverse range of real-world network topologies. Using extensive experiments, we show that ProxaDyn could achieve significantly better results compared to the state-of-the-art NDN caching strategies. © 2013 IEEE.

关键词： Data communication systems

来源：评论

学校读者我要写书评

暂无评论

Few-Shot Referring Video Single- and Multi-Object Segmentation Via Cross-Modal Affinity with Instance Sequence Matching

引用

International Journal of computer Vision 2025年 1-19页

作者： Liu, Heng Li, Guanghui Gao, Mingqi Zhen, Xiantong Zheng, Feng Wang, Yang School of Computer Science and Technology Anhui University of Technology Maxiang Road Ma’anshan243032 China Department of Computer Science and Engineering Southern University of Science and Technology Xueyuan Avenue Shenzhen518055 China School of Computer Science and Information Engineering Hefei University of Technology Feicui Road Hefei230601 China

Referring Video Object Segmentation (RVOS) aims to segment specific objects in videos based on the provided natural language descriptions. As a new supervised visual learning task, achieving RVOS for a given scene requires a substantial amount of annotated data. However, only minimal annotations are usually available for new scenes in realistic scenarios. Another practical problem is that, apart from a single object, multiple objects of the same category coexist in the same scene. Both of these issues may significantly reduce the performance of existing RVOS methods in handling real-world applications. In this paper, we propose a simple yet effective model to address these issues by incorporating a newly designed cross-modal affinity (CMA) module based on a Transformer architecture. The CMA module facilitates the establishment of multi-modal affinity over a limited number of samples, allowing the rapid acquisition of new semantic information while fostering the model’s adaptability to diverse scenarios. Furthermore, we extend our FS-RVOS approach to multiple objects through a new instance sequence matching module over CMA, which filters out all object trajectories with similarity to language features that exceed a matching threshold, thereby achieving few-shot referring multi-object segmentation (FS-RVMOS). To foster research in this field, we establish a new dataset based on currently available datasets, which covers many scenarios in terms of single-object and multi-object data, hence effectively simulating real-world scenes. Extensive experiments and comparative analyses underscore the exceptional performance of our proposed FS-RVOS and FS-RVMOS methods. Our method consistently outperforms existing related approaches through practical performance evaluations and robustness studies, achieving optimal performance on metrics across diverse benchmark tests. © The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer Nature 2025.

关键词： Modal analysis

来源：评论

学校读者我要写书评

暂无评论

Improved security with novel M-Log chaos steganography algorithm for huffman compressed english text

引用

Multimedia Tools and Applications 2025年第2期84卷 665-688页

作者： Raghuvanshi, Devvrat Joshi, Kamaldeep Nandal, Rainu Sehrawat, Harkesh Singh, Surender Singh, Srishti Department of Computer Science and Engineering University Institute of Engineering and Technology Maharshi Dayanand University Rohtak Haryana India

This paper addresses the underexplored landscape of chaotic functions in steganography, existing literature when examined under PRISMA-ScR framework it was realized that most of the studies predominantly focuses on utilizing chaotic functions for either concealing or diffusing messages within cover images. This study explored a different direction of using chaotic function as guards for concealed data by proposing the M-Log chaos algorithm which employs logarithmic chaotic function to drive a unique bomb generator equation, strategically placing bomb bits in the image. These bombs act as sentinels, safeguarding the concealed message. Successful retrieval involves diffusing the bombs before extracting data from the rest of the image. Failure to defuse these bombs prior to access results in irreversible damage to concealed information, thwarting unauthorized access. Our proposed method demonstrates notable advancements across quantitative performance metrics, including a 40% increase in payload capacity, a mean PSNR of 45.8981, and a mean SSIM of 0.9931. Future avenues include exploring multi-dimensional chaotic functions and employing nature-based searching algorithms such as genetic algorithms or particle swarm optimization for optimal chaos bomb placement. © The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer Nature 2024.

关键词： Steganography

来源：评论

学校读者我要写书评

暂无评论

Halo reduction multi-exposure image fusion technique

引用

Multimedia Tools and Applications 2025年第13期84卷 12347-12370页

作者： Sharif, Rizwan Amin, Benish Sukhia, Komal Nain Department of Electrical Engineering Institute of Space Technology Islamabad Pakistan Department of Computer Science Institute of Space Technology Islamabad Pakistan

Multi-exposure image fusion (MEF) involves combining images captured at different exposure levels to create a single, well-exposed fused image. MEF has a wide range of applications, including low light, low contrast, night photography, medical imaging, and remote sensing. However, MEF methods often face issues like artifacts, halos around edges, color inconsistencies, noise amplification, and difficulty in preserving fine details. Moreover, assessing the quality of fused images objectively is complex due to the subjective nature of human perception. Solving the challenges is essential to developing efficient MEF techniques that produce high-quality results across various scenarios. The proposed technique introduces an approach to handling halo artifacts and implementing MEF. The Dense Scale-Invariant Feature Transform (DSIFT) is used to capture vital information about image brightness, texture, and edges from source images. Three weight maps are computed from the local mean, signal strength, and the global gradient for initial weight estimation. The local mean represents the brightness of specific image areas, signal strength preserves essential details like textures and edges while reducing image noise, and global gradient helps identify regions with significant pixel value shifts across multiple exposure images. The weight maps are then combined using a weighted average and a Gaussian smoothing filter is applied to reduce inherent noise and discontinuities in the original weights. Subsequently, pyramid decomposition is performed to generate a fused image. The efficiency of the proposed method is extensively tested on challenging multi-exposure image sequences. The results of the proposed approach demonstrate its superiority in both subjective evaluation and objective metrics like the MEF-Structural Similarity Index (MEF-SSIM), Natural Image Quality Evaluator (NIQE) and Gradient based performance measure (QAB/f). © The Author(s), under exclusive licence to Springer

关键词： Image fusion

来源：评论

学校读者我要写书评

暂无评论

Deep learning and machine learning based highly accurate reflection prediction model for multi layers anti-reflection coatings

引用

Optical and Quantum Electronics 2025年第1期57卷 1-22页

作者： Oktay, Semih Duru, İremnur Bakır, Halit Tabaru, Timuçin Emre Department of Electrical-Electronics Engineering Sivas University of Science and Technology Sivas Turkey Computer Engineering Department Sivas University of Science and Technology Sivas Turkey

This study focuses on creating an accurate reflection prediction model that will guide the design of filters with multilayer Anti-Reflection Coating (ARC) to optimize the thickness parameters using Machine Learning (ML) and Deep Learning (DL) techniques. This model aims to shed light on the design process of a multilayer optical filter, making it more cost-effective by providing faster and more precise production. In creating this model, a dataset containing data obtained from 3000 (1500 Ge–Al2O3, 1500 Ge–SiO2) simulations previously performed on a computer based on the thicknesses of multilayer structural materials was used. The data are generated using Computational Electromagnetic simulation software based on the Finite-Difference Time-Domain method. To understand the mechanism of the proposed model, two different two-layer coating simulations were studied. While Ge was used as the substrate in both coatings, Al2O3 and SiO2 were used as the second layers. The data set consists of the 3–5 µm and 8–12 µm bands typical for the mid-wave infrared (MWIR) and long-wave infrared (LWIR) bands and includes reflectance values for wavelengths ranging between these spectra. In the specified 2-layer data set, the average reflectance was obtained with a minimum of 0.36 at 515 nm Ge and 910 nm SiO2 thicknesses. This value can be increased by adapting the proposed model to more than 2 layers. Six ML algorithms and a DL model, including artificial neural networks and convolutional neural networks, are evaluated to determine the most effective approach for predicting reflectance properties. Furthermore, in the proposed model, a hyperparameter tuning phase is used in the study to compare the efficiency of ML and DL methods to generate dual-band ARC and maximize the prediction accuracy of the DL algorithm. To our knowledge, this is the first time this has been implemented in this field. The results show that ML models, particularly decision tree (MSE: 0.00000069, RMSE: 0.00083), rand

关键词： Antireflection coatings

来源：评论

学校读者我要写书评

暂无评论

Optimized graph transformer with molecule attention network based multi class attack detection framework for enhancing privacy and security in WSN

引用

Multimedia Tools and Applications 2025年第15期84卷 14273-14304页

作者： Anoop, M. Mary, L. William Wilson, Allan J. Kiran, W.S. Department of Computer Science and Application SRM Institute of Science and Technology Ramapuram campus Chennai600089 India Department of Electronics and Communication Engineering Amrita College of Engineering and Technology Nagercoil India

In today's rapidly evolving network landscape, cybersecurity has become increasingly crucial. However, wireless sensor networks face unique challenges due to their limited resources and diverse composition, high computational demands, diminished accuracy rates, making them susceptible to a wide array of cyber threats. To tackle these challenges, this research introduced a novel method known as Optimized Graph Transformer with Molecule Attention Network (OGTMAN) on the fusion of Secure Multi-party Computation (SMC) and differential privacy aimed at enhancing privacy and security. This approach starts with data preprocessing, employing min–max normalization to address dataset imbalances. N-Tuple Contrastive Learning with Generative transformer(N-TCLG)feature extractor distinguish between different classes of data instances and effectively learns representations that capture relevant information about network traffic patterns Further, feature selection by enhanced chi-square statistics with kepler algorithm, is utilized to assess the significance of each feature in characteristic between normal network behavior and malicious activities further refines the feature selection method by optimizing the selection criteria based on the statistical properties of the dataset. Finally, the proposed OGTMAN classifies various attacks based on their behavioural patterns with the help of enhanced ant aphid optimization with adaptive step factor. Ultimately, this fine-tuning yield superior model training outcomes, resulting more accurate attack classification outcomes. To evaluate the effectiveness of this approach, datasets encompassing with KDD Cup 99 datasets are employed. The efficiency of the introduced method is demonstrated when compared with prevailing attack detection methods in terms of accuracy, recall, precision and F-measure of 99.98%, 99.5%.,99.9% and 99.5% respectively. © The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer N

关键词： Molecules

来源：评论

学校读者我要写书评

暂无评论

Spatial Performance Analysis and Design Principles of Random Access-Based Wireless Mutual Broadcast in Various Spectral Environments

引用

IEEE Wireless Communications Letters 2025年第5期14卷 1296-1300页

作者： Kwon, Taesoo Seoul National University of Science and Technology Department of Computer Science and Engineering Seoul01811 Korea Republic of

Proximity-aware services are expected to grow further in the B5G and 6G era, often leveraging random access based wireless mutual broadcast operations among nearby nodes. This letter establishes a tractable framework for RA-WMB performance across sub-6GHz, mmWave, and THz bands, incorporating LoS/NLoS channels, beamforming, and molecular absorption. It analytically examines the impact of beamforming on the number of successful messages and broadcast coverage, derives an explicit suboptimal transmission probability for performance maximization, and highlights the critical role of sidelobe gain through both analytical and numerical results. © 2012 IEEE.

关键词： Beamforming

来源：评论

学校读者我要写书评

暂无评论

Integrating blockchain and machine learning for enhanced anti-money laundering system

引用

International Journal of Information technology (Singapore) 2025年第4期17卷 2439-2447页

作者： Shafin, Khandakar Md Reno, Saha Department of Computer Science and Engineering CCN University of Science and Technology Cumilla 3503 Bangladesh Department of Computer Science and Engineering Ahsanullah University of Science and Technology (AUST) Dhaka 1208 Bangladesh

Money laundering is a serious threat to global financial systems, causing instability and inflation, and especially hurting middle-class savings. This paper suggests a new way to tackle these problems by using blockchain technology and advanced machine learning models. We use hyperledger fabric to securely record transactions and advanced algorithms like autoencoders and neural networks to create a strong anti-money laundering (AML) system. This system can detect and predict illegal financial activities in real-time and includes continuous monitoring and alerts. These features improve financial transparency and stability while protecting the savings of the middle class. By identifying and reducing risks early, our solution not only ensures compliance with regulations but also strengthens the resilience of global financial systems against new threats. This research helps develop effective ways to fight financial crime, promoting a safer and more transparent financial world. © Bharati Vidyapeeth's Institute of computer Applications and Management 2024.

关键词： Blockchain Hyperledger fabric Inflation Machine learning Money laundering Security

来源：评论

学校读者我要写书评

暂无评论

Video forgery localization using inter-frame denoising and intra-frame segmentation

引用

Multimedia Tools and Applications 2025年 1-17页

作者： Banerjee, Debanik Chittaragi, Nagaratna B. Koolagudi, Shashidhar G. Department of Computer Science and Engineering National Institute of Technology Karnataka Karnataka Surathkal575025 India Department of Information Science and Engineering Siddaganaga Institute of Technology Karnataka Tumkur572103 India

Video forgery detection has been necessary with recent spurt in fake videos like Deepfakes and doctored videos from multiple video capturing devices. In this paper, we provide a novel technique of detecting fake videos by creating an ensemble network, based on statistical and deep learning methods to detect interframe forgery and intraframe forgery in forged videos separately. In this paper, Noise signature extraction of a particular image capturing sensor and an Autoencoder-based Convolutional Neural Network model (CNN) are used to localize the forged regions. We have trained the model to localize Deepfake video forgeries as well as copy-paste forgeries with effective results in the test data. The proposed fake video detector can be applied at the back-end of on-line video aggregating services and check their authenticity to verify the genuineness of videos. The results achieved have shown better performances in detecting fake videos compared to existing methods. © The Author(s), under exclusive licence to Springer science+Business Media, LLC, part of Springer Nature 2025.

关键词： Convolutional neural networks

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：