检索结果-内蒙古大学图书馆

Object Segmentationin 3D Reconstructed Scenes Based on Gaussian Splatting

Jisuanji Xuebao/Chinese Journal of computers 2025年第5期48卷 1232-1243页

作者： Wang, Feng Yin, Ying Wang, Jia-Yan Tang, Yong Li, Sheng Zhao, Jing School ofInformation Science and Engineering Yanshan University Hebei Qinhuangdao066004 China Hebei Key Laboratory of Computer Virtual Technology and SystemIntegration Hebei Qinhuangdao066004 China School of Computer Science Peking University Beijing100871 China Beijing Engineering Technology Research Center of Virtual Simulationand Visualization Peking University Beijing100871 China

In the field of 3D reconstruction- the recent introduction of 3D Gaussian splatting has led to substantial advancements in both training efficiency and reconstruction quality, especially when compared to traditional neural radiance field-based approaches. These advancements make 3D Gaussian splatting a promising tool for reconstructing complex scenes with high accuracy and efficiency, significantly reducing the training time compared to the methods based on Neural Radiance Fields. However* challenges remain in effectively segmenting the reconstructed scenes* particularly when dealing with intricate occlusions and overlapping objects. Scene segmentation in such complcx environments becomes difficult due to the dynamic interactions between objects, subtle boundarics, and the presence of noise in the reconstruction process. These challenges hin-der the accurate identification and isolation of specific objects within the scene, which is critical for a ränge of practical applications. To address this issue, this papcr proposcs a novel, light-wcight method for 3D Gaussian-based scene segmentation that effectively Segments specific objects from a rcconstructed scene. The core innovation of this approach is the use of the differenti-able nature of 3D Gaussian splatting during the reconstruction process, which allows for the sim-ultaneous training of object dass features alongside the scene reconstruction. Specifically, a dass feature is associated with each Gaussian kernel, encoding semantic information that is crucial for object recognition and Classification within the 3D space. These dass features arc traincd to dis-tinguish between different objects and map them to specific object categories. After embcdding the dass features into the scene, a multi-layer perceptron decoder is designed and trained to map the dass features to object dass labels, allowing the System to identify and classify objects based on the features associated with each Gaussian kernel. During the ren

关键词： 3D reconstruction

来源：评论

学校读者我要写书评

暂无评论

Motor Fault Diagnosis Based on Short-time Fourier Transform and Convolutional Neural Network

引用

Chinese Journal of Mechanical Engineering 2017年第6期30卷 1357-1368页

作者： Li-Hua Wang Xiao-Ping Zhao Jia-Xin Wu Yang-Yang Xie Yong-Hong Zhang School of Information and Control Nanjing University ofInformation Science and Technology Nanjing 210044 China School of Computer and Software Nanjing University ofInformation Science and Technology Nanjing 210044 China Network Monitoring Center of Jiangsu Province NanjingUniversity of Information Science and TechnologyNanjing 210044 China

With the rapid development of mechanical equipment, the mechanical health monitoring field has entered the era of big data. However, the method of manual feature extraction has the disadvantages of low efficiency and poor accuracy, when handling big data. In this study, the research object was the asynchronous motor in the drivetrain diagnostics simulator system. The vibration signals of different fault motors were collected. The raw signal was pretreated using short time Fourier transform （STFT） to obtain the corresponding time-frequency map. Then, the feature of the time-frequency map was adap- tively extracted by using a convolutional neural network （CNN）. The effects of the pretreatment method, and the hyper parameters of network diagnostic accuracy, were investigated experimentally. The experimental results showed that the influence of the preprocessing method is small, and that the batch-size is the main factor affecting accuracy and training efficiency. By investigating feature visualization, it was shown that, in the case of big data, the extracted CNN features can represent complex mapping relationships between signal and health status, and can also overcome the prior knowledge and engineering experience requirement for feature extraction, which is used by tra- ditional diagnosis methods. This paper proposes a new method, based on STFT and CNN, which can complete motor fault diagnosis tasks more intelligently and accurately.

关键词： Big data Deep learning Short-time Fouriertransform Convolutional neural network Motor

来源：评论

学校读者我要写书评

暂无评论

A snoring detector for OSAHS based on patient's individual personality

A snoring detector for OSAHS based on patient's individual p...

引用

International Conference on Awareness science and technology (iCAST)

作者： Yuxia Zhao Haixiu Zhang Wenlong Liu Shuxue Ding School ofInformation and Communication Engineering Dalian University of Technology Dalian China School of Computer Science and Engineering University of Aizu Aizu-Wakamatsu Japan

ISBN: (纸本)9781457708879

A conventional diagnostic tool for assessing Obstructive Sleep Apnea Hypopnea Syndrome (OSAHS) is polysomnography (PSG), which is expensive and uncomfortable for patients. It is an important and urgent topic to find a non-invasive and low-cost diagnostic approach for OSAHS detection. Recently, the snore signal analysis receives much attention due to its potential capability for OSAHS detection. In this paper, we propose a novel method for diagnosing OSAHS based on patient's individual personality. First, the first formant frequencies of each snorer are classified into two clusters by K-means clustering. And then, using the first cluster center of each snorer, we set a personalized threshold to distinguish the hypopneic snores from the normal ones. Since the proposed threshold varies with each individual, the patient's individual personality can be overcome effectively. Experimental results show the validity of the proposed detector. In the experiments, the sensitivity of our method can achieve 90% and the specificity can achieve 91.67%.

关键词： Atmospheric modeling Argon High definition video Computational modeling Frequency estimation Educational institutions

来源：评论

学校读者我要写书评

暂无评论

Multi-focus color image fusionusing block variance voting

Multi-focus color image fusionusing block variance voting

引用

International Conference on Communication Software and Networks, ICCSN

作者： Feiran Jie Taisong Jin Lingling Li Zhenhan Lv Science and technology on electro-optic control laboratory Luoyang China Department of Computer Science and Application Zhengzhou Institute of Aeronautical Industry Management Zhengzhou China School ofInformation Science & Technology Xiamen University Xiamen China

For the multi-focus image, a method to distinguish the focal area and the blur pixels is proposed. The variance of the neighbor pixels of each pixel is calculated and different levels of variance with different weights are denoted where a pixel is in the focal area is estimated by the variance of the neighbor pixels' voting. For the binary matrix, some isolated 0 or 1 to reduce noise is eliminated. Finally, the fusion is done by the pixels of the focal area of two images. Experimental results show that the proposed method is fast and effective.

关键词： Educational institutions Laboratories Wavelet transforms

来源：评论

学校读者我要写书评

暂无评论

OpenCV 2.0 (1);outline of Open CV and how to install

引用

Kyokai Joho Imeji Zasshi/Journal of the Institute of Image Information and Television Engineers 2010年第7期64卷 987-990页

作者： Ueda, Etsuko Koeda, Masanao Suenaga, Tsuyoshi Ido, Junichi Nara National College of Technology Department of Control Engineering Nara Japan Osaka Electro-Communication University Department of Computer Sceience Osaka Japan Nara Institute of Science and Technology Graduate School OfInformation Science Nara Japan Kawada Industries Inc Tochigi Japan

来源：评论

学校读者我要写书评

暂无评论

Packet-analysis based flow identification system for Peer-to-Peer application

Packet-analysis based flow identification system for Peer-to...

引用

2009 International Conference on Test and Measurement, ICTM 2009

作者： Yu, Baolong Xiong, Ke Du, Ye Ding, Xiaoming Zhang, Xiong School of Computer and Information Technology Beijing Jiaotong University Beijing 100044 China Institute OfInformation Science Beijing Jiaotong University Beijing 100044 China

ISBN: (纸本)9781424447008

Recently, Peer-to-Peer (P2P) streaming occupies a large number of network bandwidth. To keep network bandwidth from being occupied by a certain kind of network service, the traffic generated by the service should be limited. The limitation operation often closely depends on flow identification. This paper presents a P2P flow identification method on the basis of deep-packet analysis. The proposed method uses a distributed architecture, through client-side analysis of the statistical traffic characteristics, avoiding large amounts of analysis about packet payload on the network and parsing and restoring of the protocol. It also can be used to restrict the network bandwidth occupied by a large number of traffic on the server of the export gateway, thereby reducing the non-equilibrium bandwidth occupation. The verification experiments on the overall design showthat P2P stream packets can be identified effectively by this system with more than 80% success rate. © 2009 IEEE.

关键词： Network architecture

来源：评论

学校读者我要写书评

暂无评论

Harmonious human-computer interaction in pervasive environment

Harmonious human-computer interaction in pervasive environme...

引用

2008 3rd International Conference on Pervasive Computing and Applications, ICPCA08

作者： Hong, Liu Wenhao, Li Jianguo, Li School of Information Science and Engineering Shandong Normal University Jinan China School of Computer Science and Technology Shandong University Jinan China Center ofInformation Technology Shandong Normal University Jinan China

ISBN: (纸本)9781424420209

This paper presents harmonious human-computer interaction research in pervasive environment based on dynamic sharing niche technology and co-evolutionary learning. It begins with a discussion ofsome important issues related to user-centered human-computer interaction. Then it describes niched co-evolving approach, and based on dynamic sharing niche technology and homophily principle, the group model and classified algorithm is introduced At last, itfurther explores the co-evolutional learning for increasing the self-adaptation of harmonious human-computer interaction and discusses directions forfuture research. © 2008 IEEE.

关键词： Human computer interaction

来源：评论

学校读者我要写书评

暂无评论

Collaborative research project on ultra-low-power data-driven networking system

Collaborative research project on ultra-low-power data-drive...

引用

2008 International Conference on Parallel and Distributed Processing Techniques and Applications, PDPTA 2008

作者： Nishikawa, Hiroaki Ishii, Hiroshi Iwata, Makoto Dept. of Computer Science Graduate School of Systems and Information Engineering University of Tsukuba Tsukuba Ibaraki Japan Professional Graduate School of Embedded Technology Tokai University Minato Tokyo Japan Dept. Ofinformation Systems Engineering Kochi University of Technology Kami Kochi Japan

ISBN: (纸本)1601320841

Power consumption will be one of the most crucial issues to realize both future VLSI chips and communication environments. This research project proposed in this paper is aiming at developing a data-driven networking system by which the coming networking environment can be achieved with ultra- low-power consumption 1/100 to 1/1000 of the present. To realize the objectives, this project is planning to proceed from two approaches by ultimately utilizing passive data-driven principle. One approach is from the processor platform, that is, data-driven chip multi-processor platform based on self-timed elastic pipeline. And the other is from the networking scheme, that is, ad hoc networking scheme. As for the platform approach, based on rich experience so far on data-driven processor, we propose a novel architecture including chip multi-processor core with highly effective self-timed elastic pipeline. As for the networking scheme, this paper shows our proposal can reduce the number of control packets that results in reduction of power consumption.

关键词： Electric power utilization

来源：评论

学校读者我要写书评

暂无评论

A New Alternative Blind Super-Resolution Algorithm Based on Motion Estimation

A New Alternative Blind Super-Resolution Algorithm Based on ...

引用

Congress on Image and Signal Processing, CISP

作者： Xiao-hua Yuan Shu-qin Li Chun-ping Liu Jing Li College of Information Shanghai Fisheries University Shanghai China Department of Computer Science Beijing Information Science and Technology University Beijing China School of ComputerScience &Technology Soochow University Soochow China College ofInformation Shanghai Fisheries University Shanghai China

In this paper, we propose a new alternative blind SR to estimate motion, blurring, and higher resolution (HR) image alternatively and iteratively in case the three are all unknown. The related ternary optimization is first divided into two alternate and iterative optimizations, in which the unitary one is about motion estimation and the dualistic one is about blurring and HR image estimation, and the dualistic one is divided into two alternate and iterative unitary optimizations again. Then in the related minimization functional, Total Variant regularization is used to conserve local discontinuous of the estimation, Half-quadratic regularization is used to solve the thus induced non-quadratic problem, and additive operator split is used to simplify the related computation. Test results at last show the validity of the proposed SR.

关键词： Motion estimation Strontium Image resolution Signal resolution Signal processing algorithms Educational institutions Aquaculture computer science Testing Spatial resolution

来源：评论

学校读者我要写书评

暂无评论

Time-of-arrival estimation for IR-UWB systems in dense multipath environment

Time-of-arrival estimation for IR-UWB systems in dense multi...

引用

IEEE International Symposium on Personal, Indoor and Mobile Radio Communications (PIMRC)

作者： Yizhi Xu Edward K. S. Au Albert K. Wong Qin Wang Wireless IC System Design Center HKUST Fok Ying Tung Graduate School Nansha China Modulation and Coding Department Institute for Infocomm Research Agency for Science Technology and Research Singapore Department ofElectronic and Computer Engineering Hong Kong University of Science and Technology Hong Kong China School ofInformation Engineering University of Science and Technology of Beijing Beijing China

In this paper, a novel time-of-arrival estimation algorithm is proposed for impulse radio-based UWB systems in dense multipath environment. In particular, a dedicated ranging sequence with good cross-correlation property is proposed and a threshold-based search-back unit is introduced, with the purpose of supporting high-precision ranging even at low signal-to-noise ratio. An approximate closed-form expression on the optimal threshold is derived based on the Rayleigh statistical analysis, and its accuracy and tightness are verified by simulations. Results show that the proposed ranging scheme achieves an excellent mean absolute error performance in the order of nanoseconds without introducing too much latency.

关键词： Time of arrival estimation Matched filters Design engineering Signal to noise ratio Statistical analysis Ultra wideband technology Bandwidth Mathematical model Closed-form solution Analytical models

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：