检索结果-内蒙古大学图书馆

From One to Many: Portable Model Construction With Independent Network Units

IEEE Internet of Things Journal 2025年

作者： Lu, Zhaocheng Zhang, Haofei Sun, Li Xia, Jiabin Ye, Jingwen Song, Mingli Zhejiang University School of Software Technology Hangzhou310013 China National University of Singapore Learning and Vision Laboratory Singapore Zhejiang University College of Computer Science and Technology Hangzhou310013 China

Artificial Intelligence of Things (AIoT) devices are highly versatile, operating in diverse environments, which necessitates local fine-tuning of deployed models to ensure compatibility with specific conditions. However, traditional fine-tuning methods often rely on cloud-based collaborative training, which is impractical for complete on-device deployment and incurs high re-tuning costs when the environment changes. To address these challenges, we propose a portable and modular approach named Portable model construction with Independent Network Units (POINT), which enables efficient customization for diverse tasks on the AIoT devices. POINT combines cloud-based model management with feature reuse to efficiently adapt pre-trained models using minimal local data. This enables offline finetuning directly on end-side devices without relying on extensive computational resources. Compared to Parameter-Efficient Fine-Tuning (PEFT) approaches like Low-Rank Adaptation (LoRA), POINT reduces the number of trainable parameters by up to 98.94%. During deployment, task-relevant models are selected from a centralized cloud repository and integrated as the backbone of the target model. Only a lightweight task-specific head is trained for downstream tasks, making POINT exceptionally lightweight and suitable for end-side customization. To mitigate performance degradation as the number of selected models increases, POINT incorporates a dynamic optimization strategy, which balances resource constraints and model performance by adaptively managing the learning process. Extensive experiments demonstrate that POINT achieves competitive results with only 100 images per category and 20 epochs of training, significantly reducing computational and resource requirements. Compared to traditional methods, POINT offers an efficient, scalable, and resource-conscious solution for deploying AI models in diverse and constrained AIoT environments. © 2014 IEEE.

关键词： Resource allocation

来源：评论

学校读者我要写书评

暂无评论

Reconfigurable Intelligent Surface-Aided Covert Communications: A Multi-User Scenario

引用

IEEE Transactions on Vehicular technology 2025年

作者： Hu, Xiaoyan Zhao, Pengze Xiao, Han Zheng, Tong-Xing Wang, Wenjie Yang, Kun Xi'an Jiaotong University School of Information and Communications Engineering Xi'an710049 China Southeast University National Mobile Communications Research Laboratory Nanjing210096 China Nanjing University State Key Laboratory of Novel Software Technology Nanjing210008 China School of Intelligent Software and Engineering Suzhou215163 China University of Essex School of Computer Science and Electronic Engineering United Kingdom

This paper explores a multi-user covert communication scenario with the assistance of a reconfigurable intelligent surface (RIS). Except a covert user monitored by the warden, there are multiple public users existing in the service area. For the purpose of evaluating the efficacy of the covert communication system, a covert constraint is derived measuring the minimum detection error probability of the warden through the Kullback-Leibler (KL) divergence. Then an optimization problem is formulated to maximize the convert rate, while ensuring the constraints of covertness, transmit power and quality of service (QoS) for public users. A two-step iterative algorithm jointly designing the active and passive beamforming, is put forth to efficiently resolve this non-convex optimization problem. Additionally, the numerical results clearly demonstrate that the proposed RIS-assisted scheme can significantly enhance the covert communication performance on the premise of guaranteeing the public communication performance of the system. © 1967-2012 IEEE.

关键词： Convex optimization

来源：评论

学校读者我要写书评

暂无评论

Tip the Scales: Achieving Balance in Adversarial Examples Across Modalities

Tip the Scales: Achieving Balance in Adversarial Examples Ac...

引用

International Conference on Acoustics, Speech, and Signal Processing (ICASSP)

作者： Zhenbo Shi Zhidong Yu Yuxuan Zhang Shuchang Wang Xiaoman Liu Wei Yang Liusheng Huang School of Computer Science and Technology University of Science and Technology of China Hefei China Suzhou Institute for Advanced Research University of Science and Technology of China Suzhou China Laboratory for Advanced Computing and Intelligence Engineering Wuxi China School of Software Engineering University of Science and Technology of China Suzhou China Hefei National Laboratory University of Science and Technology of China Hefei China

ISBN: (数字)9798350368741

ISBN: (纸本)9798350368758

In the field of multimodal learning, controlling the training of unimodal encoders from different perspectives is a primary approach to addressing Training Imbalance. However, the inherent capacity limitations of the modality affect the model’s capability. Therefore, generating adversarial examples that can achieve balanced transferability remains a challenging and perplexing problem. In this paper, we propose the InterModality Balanced Attack (MOBA) to address this problem. MOBA leverages Aggregated Modality Perturbation (AMP), which exploits the unbalanced effects of text and image perturbations to maximize the impact on the victim model. AMP capitalizes on the intrinsic feature connections between modalities during the optimization process, adjusting perturbations through Cross-Modality Discrepancy Loss to enhance attack success rates. Additionally, we devise the Transferability-Enhanced Evolution (TEE) to overcome the issue of diminished attack transferability due to model capacity limitations. TEE employs Transfer-Driven Optimization Loss to alleviate overfitting in single models, thereby enhancing the generalization ability.

关键词： Training Adaptation models Perturbation methods Architecture Buildings Signal processing Acoustics Speech processing Optimization Overfitting

来源：评论

学校读者我要写书评

暂无评论

Learnable Prompt as Pseudo-Imputation: Rethinking the Necessity of Traditional EHR Data Imputation in Downstream Clinical Prediction 25

Learnable Prompt as Pseudo-Imputation: Rethinking the Necess...

引用

Proceedings of the 31st ACM SIGKDD Conference on Knowledge Discovery and Data Mining V.1

作者： Weibin Liao Yinghao Zhu Zhongji Zhang Yuhang Wang Zixiang Wang Xu Chu Yasha Wang Liantao Ma School of Computer Science Peking University Beijing China & Key Laboratory of High Confidence Software Technologies Ministry of Education Beijing China National Engineering Research Center for Software Engineering Peking University Beijing China School of Software and Microelectronics Peking University Beijing China & Key Laboratory of High Confidence Software Technologies Ministry of Education Beijing China School of Computer Science Peking University Beijing China Center on Frontiers of Computing Studies Peking University Beijing China Key Laboratory of High Confidence Software Technologies Ministry of Education Beijing China & Peking University Information Technology Institute (Tianjin Binhai) Peking University Tianjin China National Engineering Research Center for Software Engineering Peking University Beijing China & Key Laboratory of High Confidence Software Technologies Ministry of Education Beijing China

ISBN: (纸本)9798400712456

Analyzing the health status of patients based on Electronic Health Records (EHR) is a fundamental research problem in medical informatics. The presence of extensive missing values in EHR makes it challenging for deep neural networks (DNNs) to directly model the patient's health status. Existing DNNs training protocols, including Impute-then-Regress Procedure and Jointly Optimizing of Impute-n-Regress Procedure, require the additional imputation models to reconstruction missing values. However, Impute-then-Regress Procedure introduces the risk of injecting imputed, non-real data into downstream clinical prediction tasks, resulting in power loss, biased estimation, and poorly performing models, while Jointly Optimizing of Impute-n-Regress Procedure is also difficult to generalize due to the complex optimization space and demanding data requirements. Inspired by the recent advanced literature of learnable prompt in the fields of NLP and CV, in this work, we rethought the necessity of the imputation model in downstream clinical tasks, and proposed Learnable Prompt as Pseudo-Imputation (PAI) as a new training protocol to assist EHR analysis. PAI no longer introduces any imputed data but constructs a learnable prompt to model the implicit preferences of the downstream model for missing values, resulting in a significant performance improvement for all state-of-the-arts EHR analysis models on four real-world datasets across two clinical prediction tasks. Further experimental analysis indicates that PAI exhibits higher robustness in situations of data insufficiency and high missing rates. More importantly, as a plug-and-play protocol, PAI can be easily integrated into any existing or even imperceptible future EHR analysis models. The code of this work is deployed publicly available at https://***/MrBlankness/PAI to help the research community reproduce the results and assist the EHR analysis tasks.

关键词： electronic health records

来源：评论

学校读者我要写书评

暂无评论

SeaCap: Multi-Sight Embedding and Alignment for One-Stage Image Captioner

引用

IEEE Transactions on Multimedia 2025年

作者： Wang, Bo Zhang, Zhao Zhao, Mingbo Jin, Xiaojie Xu, Mingliang Wang, Meng Hefei University of Technology School of Computer Science and Information Engineering Hefei230601 China Yunnan Key Laboratory of Software Engineering Yunan China Donghua University School of Computer Science and Engineering Shanghai201620 China ByteDance Culver CityCA90230 United States Zhengzhou University School of Computer and Artificial Intelligence Zhengzhou450001 China

Recent mainstream image captioning methods usually adopt two-stage captioners, i.e., calculating the object features of the given image by a pre-trained detector and then feeding them into a language model to generate the descriptive sentences. However, such a two-stage procedure will lead to a task-based information gap that decreases the performance of the captioners, because the object features learned from the detection task are suboptimal representations and cannot provide all the necessary information for subsequent sentence generation. Besides, the object features are usually represented by the last pooling features of the detector that lose the local details of images. In this paper, we propose a novel One-Stage Image Captioner using dynamic multi-sight embedding and alignment, called SeaCap, which directly transforms input images into descriptive sentences in one stage to eliminate the information gap. Specifically, to obtain rich features, we use the Swin Transformer to capture multi-level features, followed by a sights alignment module to alleviate the vision confusion, and then feed them into a novel dynamic multi-sight embedding module to exploit both the global structure and local texture of input images. To enhance the global modeling capacity of the visual encoder, we propose a new dual-dimensional refining module to non-locally model the interaction of the embedded features. As a result, SeaCap can obtain rich and useful information to improve the performance of the captioner. Extensive comparisons on the benchmark MS-COCO, Flickr8K and Flickr30 K datasets verified the superior performance of our method. © 1999-2012 IEEE.

关键词： Embeddings

来源：评论

学校读者我要写书评

暂无评论

MixSignGraph: A Sign Sequence is Worth Mixed Graphs of Nodes

arXiv

引用

arXiv 2025年

作者： Gan, Shiwei Yin, Yafeng Jiang, Zhiwei Wen, Hongkai Xie, Lei Lu, Sanglu State Key Laboratory for Novel Software Technology Nanjing University Nanjing210023 China Department of Computer Science University of Warwick CoventryCV4 7AL United Kingdom

—Recent advances in sign language research have benefited from CNN-based backbones, which are primarily transferred from traditional computer vision tasks (e.g., object identification, image recognition). However, these CNN-based backbones usually excel at extracting features like contours and texture, but may struggle with capturing sign-related features. In fact, sign language tasks require focusing on sign-related regions, including the collaboration between different regions (e.g., left hand region and right hand region) and the effective content in a single region. To capture such region-related features, we introduce MixSignGraph, which represents sign sequences as a group of mixed graphs and designs the following three graph modules for feature extraction, i.e., Local Sign Graph (LSG) module, Temporal Sign Graph (TSG) module and Hierarchical Sign Graph (HSG) module. Specifically, the LSG module learns the correlation of intra-frame cross-region features within one frame, i.e., focusing on spatial features. The TSG module tracks the interaction of inter-frame cross-region features among adjacent frames, i.e., focusing on temporal features. The HSG module aggregates the same-region features from different-granularity feature maps of a frame, i.e., focusing on hierarchical features. In addition, to further improve the performance of sign language tasks without gloss annotations, we propose a simple yet counter-intuitive Text-driven CTC Pre-training (TCP) method, which generates pseudo gloss labels from text labels for model pretraining. Extensive experiments conducted on current five public sign language datasets demonstrate the superior performance of the proposed model. Notably, our model surpasses the SOTA models on multiple sign language tasks across several datasets, without relying on any additional cues. Copyright © 2025, The Authors. All rights reserved.

关键词： Feature Selection

来源：评论

学校读者我要写书评

暂无评论

Joint DNN Partition and Thread Allocation Optimization for Energy-Harvesting MEC Systems

Joint DNN Partition and Thread Allocation Optimization for E...

引用

Design, Automation and Test in Europe Conference and Exhibition

作者： Yizhou Shi Liying Li Yue Zeng Peijin Cong Junlong Zhou School of Computer Science and Engineering Nanjing University of Science and Technology Nanjing China National Key Laboratory for Novel Software Technology Nanjing University Nanjing China National Mobile Communications Research Laboratory Southeast University Nanjing China Key Laboratory of Embedded System and Service Computing (Tongji University) Ministry of Education Shanghai China

ISBN: (数字)9783982674100

ISBN: (纸本)9798331534646

Deep neural networks (DNNs) have demonstrated exceptional performance, leading to diverse applications across various mobile devices (MDs). Considering factors like portability and environmental sustainability, an increasing number of MDs are adopting energy harvesting (EH) techniques for power supply. However, the computational intensity of DNNs presents significant challenges for their deployment on these resource-constrained devices. Existing approaches often employ DNN partition or offloading to mitigate the time and energy consumption associated with running DNNs on MDs. Nonetheless, existing methods frequently fall short in accurately modeling the execution time of DNNs, and do not consider to use thread allocation for further latency and energy consumption optimization. To solve these problems, we propose a dynamic DNN partition and thread allocation method to optimize the latency and energy consumption of running DNNs on EH-enabled MDs. Specifically, we first investigate the relationship between DNN inference latency and allocated threads and establish an accurate DNN latency prediction model. Based on the prediction model, a DRL-based DNN partition (DDP) algorithm is designed to find the optimal partitions for DNNs. A thread allocation (TA) algorithm is proposed to reduce the inference latency. Experimental results from our test-bed platform demonstrate that compared to four benchmarking methods, our scheme can reduce DNN inference latency and energy consumption by up to 37.3% and 38.5%.

关键词： Energy consumption Instruction sets Heuristic algorithms Artificial neural networks Predictive models Prediction algorithms Inference algorithms Partitioning algorithms Resource management Optimization

来源：评论

学校读者我要写书评

暂无评论

Diffusion-Noise-based Augmentation for Long-Tailed Remote Sensing Image Classification

引用

IEEE Transactions on Geoscience and Remote Sensing 2025年

作者： Wang, Qianqian Ye, Haibo Liang, Dong Huang, Sheng-Jun MIIT Key Laboratory of Pattern Analysis and Machine Intelligence Collaborative Innovation Center of Novel Software Technology and Industrialization College of Computer Science and Technology Nanjing university of Aeronautics and Astronautics Nanjing211106 China

Remote sensing image classification refers to the task that using algorithms to categorize satellite or aerial imagery into different land cover types. In the real world, long-tailed data distribution is commonly present in remote sensing image classification tasks, causing models to excessively favor sufficient head classes during training and degrading prediction accuracy for scarce tail classes. Although existing methods such as resampling and re-weighting can alleviate the issue of data imbalance to a certain extent, they struggle to sufficiently enhance the diversity of tail class samples. In recent years, some works have begun to use diffusion models to generate diverse samples to balance the class distribution. However, these approaches often overlook the distribution inconsistency between generated and real images, which will hinder the improvement of model performance. To tackle these challenges, this paper proposes a novel diffusion-noise-based augmentation method (DONA) with a two-stage training process. Before training, our specially designed conditional prompts are used together with the original training set to guide the diffusion model in image generation. Furthermore, we propose two strategies to effectively leverage the generated images, which are applied respectively at the end of the first training stage and during the second training stage. First, we design DiffCam-Mix to fuse the background of the generated data with the foreground of the original data, preserving the essential information of the original real images while incorporating the diversity of the generated ones. Second, we use cosine similarity to minimize the differences between the mixed data and their corresponding original data, further calibrating the distribution of different samples. Extensive experiments on three public datasets—SIRI-WHU-LT, PatternNet-LT, and RSI-CB256-LT—demonstrate the effectiveness of the proposed method. © 1980-2012 IEEE.

关键词： Optical character recognition

来源：评论

学校读者我要写书评

暂无评论

Towards Expanding Precise Timing for Collaborative Its with Deterministic Communications in 6G

Towards Expanding Precise Timing for Collaborative Its with ...

引用

IEEE Conference on Wireless Communications and Networking

作者： Hongxing Li Haotian Li Qianhan Gao Guochu Shou Yaqiong Liu Zhigang Guo Yihong Hu State Key Laboratory of Networking and Switching Technology School of Computer Science (National Pilot Software Engineering School) Beijing Key Laboratory of Network System Architecture and Convergence School of Information and Communication Engineering Beijing University of Posts and Telecommunications Beijing China

ISBN: (数字)9798350368369

ISBN: (纸本)9798350368376

The demand for precise timing continues to grow with the rapid development of smart cities, where intelligent transportation emerges as a critical application heavily dependent on accurate time distribution. Deterministic communication is anticipated to be a defining feature of nextgeneration mobile networks (6G), enabling end-to-end timecritical applications, and unlocking new possibilities for achieving end-to-end precise time synchronization as well. To cope with the challenge of delivering precise time synchronization across broader areas at a lower cost, this paper proposes a method to achieve end-to-end precise time synchronization by leveraging the deterministic characteristics of communication networks. Specially, the proposed approach adopts the software-Defined Networking (SDN) paradigm to implement deterministic networking for the seamless integration of Time-Sensitive Networking (TSN) and cellular networks. Furthermore, to mitigate the impact of wireless channel uncertainties on synchronization accuracy, we propose a network delay measurement mechanism based on the principle of time redundancy, designed to reduce end-to-end network delay variation. Experimental results validate the effectiveness of the proposed method, demonstrating its capability to constrain the end-to-end relative time error of wired and wireless converged networks to the microsecond scale, without necessitating modifications to the existing network infrastructure.

关键词： 6G mobile communication Wireless communication Cellular networks Uncertainty Accuracy Measurement uncertainty Transportation Delays Synchronization software defined networking

来源：评论

学校读者我要写书评

暂无评论

Wavelet-based Global-Local Interaction Network with Cross-Attention for Multi-View Diabetic Retinopathy Detection

arXiv

引用

arXiv 2025年

作者： Hu, Yongting Lin, Yuxin Liu, Chengliang Luo, Xiaoling Dou, Xiaoyan Xu, Qihao Xu, Yong School of Computer Science and Technology Harbin Institute of Technology Shenzhen Shenzhen China Shenzhen Key Laboratory of Visual Object Detection and Recognition Shenzhen China College of Computer Science and Software Engineering Shenzhen University Shenzhen China Ophthalmology Department Shenzhen Second People’s Hospital Shenzhen China

Multi-view diabetic retinopathy (DR) detection has recently emerged as a promising method to address the issue of incomplete lesions faced by single-view DR. However, it is still challenging due to the variable sizes and scattered locations of lesions. Furthermore, existing multi-view DR methods typically merge multiple views without considering the correlations and redundancies of lesion information across them. Therefore, we propose a novel method to overcome the challenges of difficult lesion information learning and inadequate multi-view fusion. Specifically, we introduce a two-branch network to obtain both local lesion features and their global dependencies. The high-frequency component of the wavelet transform is used to exploit lesion edge information, which is then enhanced by global semantic to facilitate difficult lesion learning. Additionally, we present a cross-view fusion module to improve multi-view fusion and reduce redundancy. Experimental results on large public datasets demonstrate the effectiveness of our method. The code is open sourced on https://***/HuYongting/WGLIN. Copyright © 2025, The Authors. All rights reserved.

关键词： Wavelet transforms

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：