检索结果-内蒙古大学图书馆

Multiple Deep-Belief-Network-Based Spectral-Spatial Classification of Hyperspectral Images

Tsinghua science and Technology 2019年第2期24卷 183-194页

作者： Atif Mughees Linmi Tao Pervasive Computing Division Tsinghua National Laboratory for Information Science and Technology Department of Computer Science and Technology Tsinghua University

A deep-learning-based feature extraction has recently been proposed for HyperSpectral Images(HSI)classification. A Deep Belief Network(DBN), as part of deep learning, has been used in HSI classification for deep and abstract feature extraction. However, DBN has to simultaneously deal with hundreds of features from the HSI hyper-cube, which results into complexity and leads to limited feature abstraction and performance in the presence of limited training data. Moreover, a dimensional-reduction-based solution to this issue results in the loss of valuable spectral information, thereby affecting classification performance. To address the issue, this paper presents a Spectral-Adaptive Segmented DBN(SAS-DBN) for spectral-spatial HSI classification that exploits the deep abstract features by segmenting the original spectral bands into small sets/groups of related spectral bands and processing each group separately by using local DBNs. Furthermore, spatial features are also incorporated by first applying hyper-segmentation on the HSI. These results improved data abstraction with reduced complexity and enhanced the performance of HSI classification. Local application of DBN-based feature extraction to each group of bands reduces the computational complexity and results in better feature extraction improving classification accuracy. In general, exploiting spectral features effectively through a segmented-DBN process and spatial features through hyper-segmentation and integration of spectral and spatial features for HSI classification has a major effect on the performance of HSI classification. Experimental evaluation of the proposed technique on well-known HSI standard data sets with different contexts and resolutions establishes the efficacy of the proposed techniques,wherein the results are comparable to several recently proposed HSI classification techniques.

关键词： hyperspectral image classification segmentation deep belief network support vector machine

来源：评论

学校读者我要写书评

暂无评论

Real-time 3D-aware Portrait Video Relighting

arXiv

引用

arXiv 2024年

作者： Cai, Ziqi Jiang, Kaiwen Chen, Shu-Yu Lai, Yu-Kun Fu, Hongbo Shi, Boxin Gao, Lin Beijing Key Laboratory of Mobile Computing and Pervasive Device Institute of Computing Technology Chinese Academy of Sciences China Beijing Jiaotong University China University of California San Diego United States Cardiff University United Kingdom City University of Hong Kong Hong Kong The Hong Kong University of Science and Technology Hong Kong University of Chinese Academy of Sciences China National Key Laboratory for Multimedia Information Processing School of Computer Science Peking University China National Engineering Research Center of Visual Technology School of Computer Science Peking University China

Synthesizing realistic videos of talking faces under custom lighting conditions and viewing angles benefits various downstream applications like video conferencing. However, most existing relighting methods are either time-consuming or unable to adjust the viewpoints. In this paper, we present the first real-time 3D-aware method for relighting in-the-wild videos of talking faces based on Neural Radiance Fields (NeRF). Given an input portrait video, our method can synthesize talking faces under both novel views and novel lighting conditions with a photo-realistic and disentangled 3D representation. Specifically, we infer an albedo tri-plane, as well as a shading tri-plane based on a desired lighting condition for each video frame with fast dual-encoders. We also leverage a temporal consistency network to ensure smooth transitions and reduce flickering artifacts. Our method runs at 32.98 fps on consumer-level hardware and achieves state-of-the-art results in terms of reconstruction quality, lighting error, lighting instability, temporal consistency and inference speed. We demonstrate the effectiveness and interactivity of our method on various portrait videos with diverse lighting and viewing conditions. Copyright © 2024, The Authors. All rights reserved.

关键词： Video conferencing

来源：评论

学校读者我要写书评

暂无评论

Dynamic Spectrum Aggregation and Access for Rescue Cognitive Networks Using Multi-Agent Actor-Critic Reinforcement Learning

引用

IEEE Transactions on Cognitive Communications and Networking 2025年

作者： Zhang, Wensheng Ding, Wenjiao Wang, Cheng-Xiang Chen, Yunfei Abreu, Giuseppe Liu, Zhi Sun, Jian Han, Zhu Shandong University School of Information Science and Engineering Shandong Key Laboratory of Intelligent Communication and Sensing-Computing Integration Qingdao266237 China Industrial and Commercial Bank Of China Limited Shandong Branch Jinan China Southeast University National Mobile Communications Research Laboratory School of Information Science and Engineering Nanjing211189 China Purple Mountain Laboratories Pervasive Communication Research Center Nanjing211111 China Durham University Department of Engineering DurhamDH1 3LE United Kingdom Constructor University Bremen GGMBH School of Computer Science and Engineering Campus Ring 1 Bremen28757 Germany University of Houston Department of Electrical and Computer Engineering HoustonTX77004 United States Kyung Hee University Department of Computer Science and Engineering Seoul446-701 Korea Republic of

Unmanned aerial vehicles (UAVs) based emergency communication is gaining more attention for disaster rescue, where the UAV can work as an aerial base station to provide spectrum access in areas with power outages or network interruptions. However, UAVs can only provide limited spectrum access for a short period of time due to limited battery energy. Aiming to providing spectrum access to more users under spectrum and energy limitations, a new dynamic spectrum aggregation and access scheme is proposed for rescue cognitive networks (RCNs), in which different kinds of users are modeled as primary users (PUs) and secondary users (SUs). In each time slot, SUs with different spectrum sensing capabilities, spectrum aggregation capabilities, and bandwidth demands need to sense and aggregate multiple spectrum slots that are not occupied by PUs for data transmission. Based on selected channel states, the reward is designed as a function of transmission rates and is fed back to SUs. A maximum entropy based multi-agent actor-critic (ME-MAAC) algorithm is proposed to optimize the action selection policy for SUs. Simulation results show that the proposed ME-MAAC algorithm can effectively achieve higher achievable rate, lower mean collision rate and lower power consumption, than state-of-the-art methods based on Deep Q-Network (DQN) algorithm. The proposed RCNs provide a novel spectrum usage paradigm for emergency communication and rescue networks. © 2015 IEEE.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Reinforcement Learning-Based Sensitive Semantic Location Privacy Protection for VANETs

引用

China Communications 2021年第6期18卷 244-260页

作者： Minghui Min Weihang Wang Liang Xiao Yilin Xiao Zhu Han Department of Information and Communication Engineering Xiamen UniversityXiamen 361005China School of Information and Control Engineering ChinaUniversity of Mining and TechnologyXuzhou 221116China The Affiliated Hospital of China University of Mining and Technology Xuzhou 221116China Department of Electrical and Computer Engineering University of HoustonHouston TX 77004USA Beijing Key Laboratory of Mobile Computing and Pervasive Device No.6 Kexueyuan South RoadBeijing 100190China

Location-based services(LBS)in vehicular ad hoc networks(VANETs)must protect users’privacy and address the threat of the exposure of sensitive locations during LBS *** release not only their geographical but also semantic information of the visited places(e.g.,hospital).This sensitive information enables the inference attacker to exploit the users’preferences and life *** this paper we propose a reinforcement learning(RL)based sensitive semantic location privacy protection *** scheme uses the idea of differential privacy to randomize the released vehicle locations and adaptively selects the perturbation policy based on the sensitivity of the semantic location and the attack *** scheme enables a vehicle to optimize the perturbation policy in terms of the privacy and the quality of service(QoS)loss without being aware of the current inference attack model in a dynamic privacy protection *** solve the location protection problem with high-dimensional and continuous-valued perturbation policy variables,a deep deterministic policy gradientbased semantic location perturbation scheme(DSLP)is *** actor part is used to generate continuous privacy budget and perturbation angle,and the critic part is used to estimate the performance of the *** demonstrate the DSLP-based scheme outperforms the benchmark schemes,which increases the privacy,reduces the QoS loss,and increases the utility of the vehicle.

关键词： semantic location sensitivity locationbased services VANET differential privacy reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

AngleSizer: Enhancing Spatial Scale Perception for the Visually Impaired with an Interactive Smartphone Assistant

arXiv

引用

arXiv 2024年

作者： Jing, Xiaoqing Yu, Chun Yue, Kun Lu, Liangyou Gao, Nan Shi, Weinan Zhang, Mingshan Wang, Ruolin Shi, Yuanchun Department of Computer Science and Technology Tsinghua University Beijing China Key Laboratory of Pervasive Computing Ministry of Education Beijing China School of Software Tsinghua University Beijing China University of California Los Angeles Los AngelesCA United States Qinghai University Qinghai Province Xining China

Spatial perception, particularly at small and medium scales, is an essential human sense but poses a significant challenge for the blind and visually impaired (BVI). Traditional learning methods for BVI individuals are often constrained by the limited availability of suitable learning environments and high associated costs. To tackle these barriers, we conducted comprehensive studies to delve into the real-world challenges faced by the BVI community. We have identified several key factors hindering their spatial perception, including the high social cost of seeking assistance, inefficient methods of information intake, cognitive and behavioral disconnects, and a lack of opportunities for hands-on exploration. As a result, we developed AngleSizer, an innovative teaching assistant that leverages smartphone technology. AngleSizer is designed to enable BVI individuals to use natural interaction gestures to try, feel, understand, and learn about sizes and angles effectively. This tool incorporates dual vibration-audio feedback, carefully crafted teaching processes, and specialized learning modules to enhance the learning experience. Extensive user experiments validated its efficacy and applicability with diverse abilities and visual conditions. Ultimately, our research not only expands the understanding of BVI behavioral patterns but also greatly improves their spatial perception capabilities, in a way that is both cost-effective and allows for independent learning. Copyright © 2024, The Authors. All rights reserved.

关键词： Smartphones

来源：评论

学校读者我要写书评

暂无评论

An Efficient Spatial-Temporal Representation Method for EEG Emotion Recognition

An Efficient Spatial-Temporal Representation Method for EEG ...

引用

Symposia and Workshops on Ubiquitous, Autonomic and Trusted computing, UIC-ATC

作者： Weining Weng Yang Gu Yiqiang Chen Guoqiang Wang Nianfeng Shi Institute of Computing Technology Chinese Academy of Sciences Beijing China University of Chinese Academy of Sciences Beijing China The Beijing Key Laboratory of Mobile Computing and Pervasive Device Beijing China School of Computer and Information Engineering Luoyang Institute of Science and Technology Luoyang China

Affective computing based on electroencephalogram (EEG) has been widely studied due to the high correlation between the brain and emotions. Neurological research shows that emotions are generated by the interaction of neural circuits separated in different brain regions. Recently, many deep learning methods have been demonstrated to recognize different emotions. However, these methods extract features at the data level, ignoring the brain’s neural mechanisms of emotion generation, which may result in the features’ weak ability to express emotion. To solve this problem and extract more discriminant features according to the neural mechanism of emotion generation, this article proposes an attention based spatial-temporal representation method (ASTR) for emotion recognition. The ASTR method constructs spatial-temporal representations of emotion from neural circuits’ activity in different brain regions. In the model, the cascade attention quantifies the multi-granularity correlation and information exchange pattern between brain regions as the spatial features, which measure neural circuits’ activation and information interaction in different brain regions. The one-way transformer introduces temporal order in spatial features, calculates temporal variation features from the long-term activity of neural circuits and generates the emotion representation. The representation incorporates spatial-temporal features of neural circuits’ activity during emotion generation, which is highly related to emotion. We conducted experiments on SEED and DEAP datasets. The experimental results show that the emotion recognition accuracy of our model is about 3.0% higher than that of the SOTA algorithms, reaching more than 98%. The results also visualize the emotion generation process, revealing how the brain works under different emotions.

关键词：

来源：评论

学校读者我要写书评

暂无评论

A Comprehensive Dataset and Automated Pipeline for Nailfold Capillary Analysis

A Comprehensive Dataset and Automated Pipeline for Nailfold ...

引用

IEEE International Symposium on Biomedical Imaging

作者： Linxi Zhao Jiankai Tang Dongyu Chen Xiaohong Liu Yong Zhou Yuanchun Shi Guangyu Wang Yuntao Wang Department of Computer Science and Technology Key Laboratory of Pervasive Computing Ministry of Education Tsinghua University Beijing China UCL Cancer Institute University College London London UK Clinical Research Institute Shanghai General Hospital Shanghai China State Key Laboratory of Networking and Switching Technology Beijing University of Posts and Telecommunications Beijing China

ISBN: (数字)9798350313338

ISBN: (纸本)9798350313345

Nailfold capillaroscopy is widely used in assessing health conditions, highlighting the pressing need for an automated nailfold capillary analysis system. In this study, we present a pioneering effort in constructing a comprehensive nailfold capillary dataset—321 images, 219 videos from 68 subjects, with clinic reports and expert annotations—that serves as a crucial resource for training deep-learning models. Leveraging this dataset, we finetuned three deep learning models with expert annotations as supervised labels and integrated them into a novel end-to-end nailfold capillary analysis pipeline. This pipeline excels in automatically detecting and measuring a wide range of size factors, morphological features, and dynamic aspects of nailfold capillaries. We compared our outcomes with clinical reports. Experiment results showed that our automated pipeline achieves an average of sub-pixel level precision in measurements and 89.9% accuracy in identifying morphological abnormalities. These results underscore its potential for advancing quantitative medical research and enabling pervasive computing in healthcare. Our data and code are available at https://***/THU-CS-PI-LAB/ANFC-Automated-Nailfold-Capillary.

关键词： Training Analytical models Solid modeling Accuracy Annotations Pipelines Biomedical measurement

来源：评论

学校读者我要写书评

暂无评论

SpeechMirror: A Multimodal Visual Analytics System for Personalized Reflection of Online Public Speaking Effectiveness

arXiv

引用

arXiv 2023年

作者： Huang, Zeyuan He, Qiang Maher, Kevin Deng, Xiaoming Lai, Yu-Kun Ma, Cuixia Qin, Sheng-Feng Liu, Yong-Jin Wang, Hongan The Beijing Key Laboratory of Human-Computer Interaction Institute of Software Chinese Academy of Sciences and also with the School of Computer Science and Technology University of Chinese Academy of Sciences China Diatom Design Limited Liability Company United States The School of Computer Science and Informatics Cardiff University United Kingdom The School of Design Northumbria University United Kingdom The Department of Computer Science and Technology MOE-Key Laboratory of Pervasive Computing Tsinghua University China

As communications are increasingly taking place virtually, the ability to present well online is becoming an indispensable skill. Online speakers are facing unique challenges in engaging with remote audiences. However, there has been a lack of evidence-based analytical systems for people to comprehensively evaluate online speeches and further discover possibilities for improvement. This paper introduces SpeechMirror, a visual analytics system facilitating reflection on a speech based on insights from a collection of online speeches. The system estimates the impact of different speech techniques on effectiveness and applies them to a speech to give users awareness of the performance of speech techniques. A similarity recommendation approach based on speech factors or script content supports guided exploration to expand knowledge of presentation evidence and accelerate the discovery of speech delivery possibilities. SpeechMirror provides intuitive visualizations and interactions for users to understand speech factors. Among them, SpeechTwin, a novel multimodal visual summary of speech, supports rapid understanding of critical speech factors and comparison of different speech samples, and SpeechPlayer augments the speech video by integrating visualization of the speaker's body language with interaction, for focused analysis. The system utilizes visualizations suited to the distinct nature of different speech factors for user comprehension. The proposed system and visualization techniques were evaluated with domain experts and amateurs, demonstrating usability for users with low visualization literacy and its efficacy in assisting users to develop insights for potential improvement. Copyright © 2023, The Authors. All rights reserved.

关键词： Modal analysis

来源：评论

学校读者我要写书评

暂无评论

Optimizing Tandem Solar Cell Efficiency: A Perovskite-CIGS Approach

Optimizing Tandem Solar Cell Efficiency: A Perovskite-CIGS A...

引用

Emerging Trends in Information Technology and Engineering (ic-ETITE), Conference on

作者： Arpita Chakraborty Aditya Chakraborty Fariba Tasnia Khan Tanjim Mahmud Mohammad Shahadat Hossain Karl Andersson Dept. of CSE Chittagong University of Engineering and Technology Chattogram Bangladesh Dept. of EEE Chittagong University of Engineering and Technology Chattogram Bangladesh Computer Science and Engineering Southern University Bangladesh Chattogram Bangladesh Dept. of CSE Rangamati Science and Technology University Rangamati Bangladesh Dept. of CSE University of Chittagong Chittagong Bangladesh Pervasive and Mobile Computing Laboratory Luleå University of Technology Luleå Sweden

Tandem solar cells combining perovskite and CIGS have rapidly gained prominence as highly efficient photovoltaic devices. Recent advancements in enhancing their efficiency in-volve adjusting the thickness of distinct layers with tunable bandgaps, demonstrating superior performance compared to al-ternative tandem solar cells. The upper cell, employing perovskite material chosen for its excellent light absorption coefficient and optoelectronic properties, collaborates with the bottom CIGS cell, characterized by a tunable bandgap and low thermal requirements, contributing significantly to overall efficiency improvement. Through numerical simulations conducted with wxAMPS software, our proposed model, which incorporates a 3μm thick perovskite absorber layer as the upper cell and a 3.5μm thick CIGS absorber layer as the lower cell, demonstrated outstanding power conversion efficiency of 40.0209%, Voc of 1.8675 V, Jsc of 23.2146 mA/cm 2 , and FF of 92.3152

关键词：

来源：评论

学校读者我要写书评

暂无评论

Deep Robust Reversible Watermarking

arXiv

引用

arXiv 2025年

作者： Chen, Jiale Wang, Wei Shi, Chongyang Dong, Li Li, Yuanman Hu, Xiping School of Computer Science Beijing Institute of Technology Beijing100081 China Guangdong-Hong Kong-Macao Joint Laboratory for Emotional Intelligence and Pervasive Computing Artificial Intelligence Research Institute Shenzhen MSU-BIT University Shenzhen518172 China School of Medical Technology Beijing Institute of Technology Beijing100081 China College of Electronics and Information Engineering Shenzhen University Shenzhen518060 China Department of Computer Science Faculty of Electrical Engineering and Computer Science Ningbo University Ningbo315211 China

Robust Reversible Watermarking (RRW) enables perfect recovery of cover images and watermarks in lossless channels while ensuring robust watermark extraction under lossy channels. However, existing RRW methods, mostly non-deep learning-based, suffer from complex designs, high computational costs, and poor robustness limiting their practical applications. To address these issues, this paper proposes Deep Robust Reversible Watermarking (DRRW), a deep learning-based RRW scheme. DRRW introduces an Integer Invertible Watermark Network (iIWN) to achieve an invertible mapping between integer data distributions, fundamentally addressing the limitations of conventional RRW approaches. Unlike traditional RRW methods requiring task-specific designs for different distortions, DRRW adopts an encoder-noise layer-decoder framework, enabling adaptive robustness against various distortions through end-to-end training. During inference, the cover image and watermark are mapped into an overflowed stego image and latent variables. Arithmetic coding efficiently compresses these into a compact bitstream, which is embedded via reversible data hiding to ensure lossless recovery of both the image and watermark. To reduce pixel overflow, we introduce an overflow penalty loss, significantly shortening the auxiliary bitstream while improving both robustness and stego image quality. Additionally, we propose an adaptive weight adjustment strategy that eliminates the need to manually preset the watermark loss weight, ensuring improved training stability and performance. Experiments on multiple datasets demonstrate that DRRW achieves notable performance advantages. Compared to state-of-the-art RRW methods, DRRW improves robustness and reduces embedding, extraction, and recovery complexities by 55.14×, 5.95×, and 3.57×, respectively. The auxiliary bitstream is shortened by 43.86×, and reversible embedding succeeds on 16,762 images in the PASCAL VOC 2012 dataset, marking a significant step toward pra

关键词： Image enhancement

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：