检索结果-内蒙古大学图书馆

Collaborative image compression and classification with multi-task learning for visual Internet of Things

Chinese Journal of Aeronautics 2022年第5期35卷 390-399页

作者： Bing DU Yiping DUAN Hang ZHANG Xiaoming TAO Yue WU Congchong RU School of Computer and Communication Engineering University of Science and Technology BeijingBeijing 100083China Department of Electronic Engineering Tsinghua UniversityBeijing 100084China School of Computer Science and Technology Xidian UniversityXian 710071China Smart Network Computing Lab Cross-strait Tsinghua Research InstitutionBeijing 100084China

Widespread deployment of the Internet of Things(Io T)has changed the way that network services are developed,deployed,and *** onboard advanced Io T devices are equipped with visual sensors that form the so-called visual Io ***,the sender would compress images,and then through the communication network,the receiver would decode images,and then analyze the images for ***,image compression and semantic inference are generally conducted separately,and thus,current compression algorithms cannot be transplanted for the use of semantic inference directly.A collaborative image compression and classification framework for visual Io T applications is proposed,which combines image compression with semantic inference by using multi-task *** particular,the multi-task Generative Adversarial Networks(GANs)are described,which include encoder,quantizer,generator,discriminator,and classifier to conduct simultaneously image compression and *** key to the proposed framework is the quantized latent representation used for compression and *** with perceptual quality can achieve low bitrate compression and reduce the amount of data *** addition,the design in which two tasks share the same feature can greatly reduce computing resources,which is especially applicable for environments with extremely limited *** extensive experiments,the collaborative compression and classification framework is effective and useful for visual IoT applications.

关键词： Deep learning Generative Adversarial Network(GAN) Image classification Image compression Internet of Things

来源：评论

学校读者我要写书评

暂无评论

Multicore Software Engineering, Performance, and Tools 1

引用

丛书名： Lecture Notes in computer science

1000年

作者： Victor Pankratius Michael Philippsen

ISBN: (数字)9783642312021

ISBN: (纸本)9783642312014

This book constitutes the refereed proceedings of the International Conference on Multiscore Software Engineering, Performance, and Tools, MSEPT 2012, held in Prague in May/June 2012. The 9 revised papers, 4 of which are short papers were carefully reviewed and selected from 24 submissions. The papers address new work on optimization of multicore software, program analysis, and automatic parallelization. They also provide new perspectives on programming models as well as on applications of multicore systems.

关键词： Special Purpose and Application-Based Systems System Performance and Evaluation Software Engineering programming Languages, Compilers, Interpreters Operating Systems programming Techniques

来源：评论

学校读者我要写书评

暂无评论

Algorithmic Mirror: Designing an Interactive Tool to Promote Self-Reflection for YouTube Recommendations

arXiv

引用

arXiv 2025年

作者： Kondo, Yui Dunnell, Kevin Xiao, Qing Zhao, Jun Rocher, Luc Oxford Internet Institute University of Oxford United Kingdom MIT Media Lab Massachusetts Institute of Technology United States Human-Computer Interaction Institute Carnegie Mellon University United States Department of Computer Science University of Oxford United Kingdom

Big Data analytics and Artificial Intelligence systems derive non-intuitive and often unverifiable inferences about individuals’ behaviors, preferences, and private lives. Drawing on diverse, feature-rich datasets of unpredictable value, these systems erode the intuitive connection between our actions and how we are perceived, diminishing control over our digital identities. While Explainable Artificial Intelligence scholars have attempted to explain the inner workings of algorithms, their visualizations frequently overwhelm end-users with complexity. This research introduces ‘hypothetical inference’—a novel approach that uses language models to simulate how algorithms might interpret users’ digital footprints and infer personal characteristics without requiring access to proprietary platform algorithms. Through empirical studies with fourteen adult participants, we identified three key design opportunities to foster critical algorithmic literacy: (1) reassembling scattered digital footprints into a unified map, (2) simulating algorithmic inference through LLM-generated interpretations, and (3) incorporating temporal dimensions to visualize evolving patterns. This research lays the groundwork for tools that can help users recognize the influence of data on platforms and develop greater autonomy in increasingly algorithm-mediated digital environments. Copyright © 2025, The Authors. All rights reserved.

关键词： Video analysis

来源：评论

学校读者我要写书评

暂无评论

Recurrent Convolutional Attention Neural Model for Sentiment Classification of short text

Recurrent Convolutional Attention Neural Model for Sentiment...

引用

2019 International Conference on Cutting-Edge Technologies in Engineering, ICon-CuTE 2019

作者： Usama, Mohd Ahmad, Belal Singh, Abhay Pratap Ahmad, Parvez Embedded and Pervasive Computing Lab School of Computer Science and Technology Huazhong University of Science and Technology Wuhan China Gurukul Kangri vishwavidyalaya Department of Computer Science Haridwar India School of Computer Science and Technology Huazhong University of Science and Technology Wuhan China

ISBN: (纸本)9781728100005

Nowadays neural attention model doing good in many tasks of natural language processing (NLP). More specifically attention mechanism has been used widely with convolutional neural network (CNN) and recurrent neural network (RNN) for many task of NLP. But the sentiment classification of short text is a challenging task because it contains limited contextual information. Thus, we proposed a new recurrent convolutional attention neural model for sentiment classification of the short text by using attention mechanism with RCNN (recurrent convolutional neural network). In proposed model attention score is calculated by averaging hidden units (feature maps) generated from LSTM (long short-term memory). Then we combined this attention score with recurrent convolution-based encoded text features to get final sentence representation. Here attention will focus on important text features and recurrent convolution makes full use of limited contextual information by processing sentence representation through different windows sizes with specialized recurrent convolution operation. Validation of the proposed model is done through experimentation with three benchmark datasets i.e. MR, SSTl, and SST2. Achieved results exhibit that our model performs better than many existing baselines works on all three datasets. © 2019 IEEE.

关键词： Convolution

来源：评论

学校读者我要写书评

暂无评论

Age of Semantics in Cooperative Communications: To Expedite Simulation Towards Real via Offline Reinforcement Learning

引用

IEEE Transactions on Network science and Engineering 2025年

作者： Chen, Xianfu Zhao, Zhifeng Mao, Shiwen Wu, Celimuge Zhang, Honggang Bennis, Mehdi Shenzhen CyberAray Network Technology Co. Ltd. Shenzhen518100 China Zhejiang University Zhejiang Lab Hangzhou310027 China Zhejiang University College of Information Science and Electronic Engineering Hangzhou310027 China Auburn University Department of Electrical and Computer Engineering AuburnAL36849 United States The University of Electro-Communications Meta-Networking Research Center Chofu-shi Tokyo182-8585 Japan The University of Electro-Communications Graduate School of Informatics and Engineering Chofu-shi Tokyo182-8585 Japan City University of Macau Faculty of Data Science Taipa999078 China University of Oulu Centre for Wireless Communications Oulu90570 Finland

The age of information metric fails to correctly describe the intrinsic semantics of a status update. In an intelligent reflecting surface-aided cooperative relay communication system, we propose the age of semantics (AoS) for measuring the semantics freshness of status updates. Specifically, we focus on status updates from a source node (SN) to the destination, which is formulated as a Markov decision process. The objective of the SN is to maximize the expected satisfaction of AoS and energy consumption under the maximum transmit power constraint. To seek the optimal control policy, we first derive an online deep actor-critic (DAC) learning scheme under the on-policy temporal difference learning framework. However, implementing the online DAC in practice poses key challenge in infinitely repeated interactions between the SN and the system, which can be dangerous particularly during exploration. We then put forward a novel offline DAC scheme, which estimates the optimal control policy from a previously collected dataset without any further interactions with the system. Numerical experiments verify the theoretical results and show that our offline DAC scheme significantly outperforms the online DAC scheme and the most representative baselines in terms of mean utility, demonstrating strong robustness to dataset quality. © 2013 IEEE.

关键词： Deep reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Intra-class patch swap for self-distillation

引用

Neurocomputing 2025年 645卷

作者： Choi, Hongjun Jeon, Eun Som Shukla, Ankita Turaga, Pavan Lawrence Livermore National Laboratory LivermoreCA94550 United States Department of Computer Science and Engineering Seoul National University of Science and Technology Seoul01811 Korea Republic of Department of Computer Science and Engineering University of Nevada RenoNV89557 United States Geometric Media Lab School of Arts Media and Engineering and School of Electrical Computer and Energy Engineering Arizona State University TempeAZ85281 United States

Knowledge distillation (KD) is a valuable technique for compressing large deep learning models into smaller, edge-suitable networks. However, conventional KD frameworks rely on pre-trained high-capacity teacher networks, which introduce significant challenges such as increased memory/storage requirements, additional training costs, and ambiguity in selecting an appropriate teacher for a given student model. Although a teacher-free distillation (self-distillation) has emerged as a promising alternative, many existing approaches still rely on architectural modifications or complex training procedures, which limit their generality and efficiency. To address these limitations, we propose a novel framework based on teacher-free distillation that operates using a single student network without any auxiliary components, architectural modifications, or additional learnable parameters. Our approach is built on a simple yet highly effective augmentation, called intra-class patch swap augmentation. This augmentation simulates a teacher–student dynamic within a single model by generating pairs of intra-class samples with varying confidence levels, and then applying instance-to-instance distillation to align their predictive distributions. Our method is conceptually simple, model-agnostic, and easy to implement, requiring only a single augmentation function. Extensive experiments across image classification, semantic segmentation, and object detection show that our method consistently outperforms both existing self-distillation baselines and conventional teacher-based KD approaches. These results suggest that the success of self-distillation could hinge on the design of the augmentation itself. Our codes are available at https://***/hchoi71/Intra-class-Patch-Swap. © 2025 Elsevier B.V.

关键词： Semantic Segmentation

来源：评论

学校读者我要写书评

暂无评论

High-speed migration by anticipative mobility

Lecture Notes in Computer Science (including subseries Lectu...

引用

Lecture Notes in computer science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) 2003年 2913卷 500-510页

作者： Stoops, Luk Verelst, Karsten Mens, Tom D'Hondt, Theo Department of Computer Science Programming Technology Laboratory Vrije Universiteit Brussel Belgium

In the advent of ambient intelligence, introducing ubiquitous mobile systems and services in general and mobile code in particular, network latency becomes a critical factor, especially in wireless, low-bandwidth environments. This paper investigates anticipative mobility, a high performance computing technique that exploits parallelism between loading and execution of applications to reduce network latency. The technique sends the mobile application code anticipative to the remote host, long before the actual migration is requested. Then, at the actual migration, we won't transfer the complete application anymore but only the delta of the current computational state with the already migrated computational state. Our experiments show that some applications can migrate in 2% of their original migration time. This allows applications to migrate very fast from host to host without a significant loss of execution time during the migration phase. © Springer-Verlag Berlin Heidelberg 2003.

关键词： Ambient intelligence

来源：评论

学校读者我要写书评

暂无评论

A Novel Twitter Spam Detection Technique by Integrating Inception Network with Attention based LSTM 5

A Novel Twitter Spam Detection Technique by Integrating Ince...

引用

5th International Conference on Trends in Electronics and Informatics, ICOEI 2021

作者： Neha, M.V. Nair, Madhu S. Artificial Intelligence Computer Vision Lab Cochin University of Science and Technology Department of Computer Science Kerala Kochi682022 India

ISBN: (纸本)9781665415712

Online Social Networking sites have become a well-known way for web surfers to connect and meet. Twitter got to be a well-known micro blogging site that clients post and associate with messages known as tweets. As this networking site gains its popularity, spammers target Twitter to spread spam posts. Hence, several spam detection techniques have been proposed by analysts to create Twitter a spam-free stage. Be that as it may, the accessible machine learning algorithms cannot effectively distin- guish spammers on Twitter because of reasonable information controls by unsolicited clients to elude spam discovery. As a result, here, we present an incipient approach predicated on a deep learning technique that leverages a text-predicated feature to detect spammers. A novel architecture that contains a one-dimensional dimension reduction inception module stacked with LSTM along with an attention layer is introduced here. Within the proposed model, the inception module extricates the features from the vectors after GloVe word embedding, and then LSTM is utilized to get the context representations. An Attention layer is also used in this model to give attention to the data outputted from LSTM module. At long last, the sigmoid classifier is utilized to classify the labels as spam or ham. Here, the execution of our proposed model is compared with four machine learning-based and two deep learning-based approaches, exhibiting our approach acquired the best results with an F1-score of 95.74, accuracy of 95.75, and precision of 95.58. © 2021 IEEE.

关键词： Feature extraction

来源：评论

学校读者我要写书评

暂无评论

Heart sounds monitor and analysis in noisy environments

Heart sounds monitor and analysis in noisy environments

引用

2012 International Conference on Systems and Informatics, ICSAI 2012

作者： Song, Dandan Jia, Lijuan Lu, Yao Tao, Linmi Beijing Lab of Intelligent Information Technology School of Computer Science Beijing Institute of Technology Beijing China Department of Computer Science and Technology Tsinghua University Beijing China

ISBN: (纸本)9781467301992

Changes and murmurs of heart sounds often provide the earliest information of abnormal heart status, which makes heart sound monitor and diagnose an important mean for cardiovascular disease. Compared to other heart monitoring approaches, such as EGG, Color Doppler and Magnetic Resonance Imaging Diagnosis, heart sound detection is more portable, much safer and cheaper, and needs lower professional requirements, thus has a broader applicability. This paper aims to design a system for monitoring and analyzing heart sounds in noisy environments, which can achieve signal acquisition, automatic analysis and classification. For the signal acquisition part, a novel equipment prototype is designed by separately using the two channels of a digital sound, one to record mixed signals of heart sounds and some environmental noises, and the other to record only environmental noises as a reference, then the environmental noise can be filtered out based on Least Mean Square (LMS) adaptive noise cancellation to get purified heart sound. By disassembling and assembling a stethoscope and a recorder, this design is realized and verified. For the signal analysis and classification part, with the features extracted by wavelet decomposition and Normalized Average Shannon Energy, a novel fuzzy neural network method with structure learning is used for the heart sound classification. Experiments with real and simulated data demonstrated that our approach can correctly classify all the tested heart sounds with noises even for the ones with previous unseen heart diseases. © 2012 IEEE.

关键词： Diagnosis

来源：评论

学校读者我要写书评

暂无评论

BlindType: Eyes-free text entry on handheld touchpad by leveraging thumb's muscle memory

引用

Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies 2017年第2期1卷 1–24页

作者： Lu, Yiqin Yu, Chun Yi, Xin Shi, Yuanchun Zhao, Shengdong Department of Computer Science and Technology Tsinghua University Tsinghua National Laboratory for Information Science and Technology Beijing Key Lab of Networked Multimedia China National University of Singapore Singapore

Eyes-free input is desirable for ubiquitous computing, since interacting with mobile and wearable devices often competes for visual attention with other devices and tasks. In this paper, we explore eyes-free typing on a touchpad using one thumb, wherein a user taps on an imaginary QWERTY keyboard while receiving text feedback on a separate screen. Our hypothesis is that users can transfer their typing ability obtained from visible keyboards to eyes-free use. We propose two statistical decoding algorithms to infer users' eyes-free input: the absolute algorithm and the relative algorithm. The absolute algorithm infers user input based on the absolute position of touch endpoints, while the relative algorithm infers based on the vectors between successive touch endpoints. Evaluation results showed users could achieve satisfying performance with both algorithms. Text entry rate was 17-23 WPM (words per minute) depending on the algorithm used. In comparison, a baseline cursor-based text entry method yielded only 7.66 WPM. In conclusion, our research demonstrates for the first time the feasibility of thumb-based eyes-free typing, which provides a new possibility for text entry on ubiquitous computing platforms such as smart TVs and HMDs. © 2017 Association for Computing Machinery.

关键词： Ubiquitous 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：