检索结果-内蒙古大学图书馆

6th IEEE International Conference on Computer and Communications, ICCC 2020

作者： Jie, Zhu Muqing, Wu Weiyao, Xu Beijing University of Posts and Telecommunications School of Information and Communication Engineering Beijing Laboratory of Advanced Information Networks Beijing Key Laboratory of Network System Architecture and Convergence Acknowledgment Beijing China

ISBN: (纸本)9781728186351

In order to solve the problem of imperfect capture of visual rhythm in action recognition, this paper proposes a novel model which is a combination of a two-pathway network and temporal pyramid networks (TPNs). Specifically, our work involves two aspects, on the one hand, we integrate TPNs into the fast pathway and the slow pathway of SlowFast network to capture multi-level features, and then merge the prediction results of the two pathways in the final recognition stage, which boosts performance of our network by enhancing the semantics extraction at input layer and feature layer. On the other hand, we apply a ConvLSTM module to improve the capability of temporal modeling in TPN, which can further strengthen the capture of features in the long-term dimensions, and the advanced TPN promotes the fusion of temporal and spatial features. Experiments on the Kinetics-400 dataset demonstrate the superiority of our novel architecture combining two-pathway network and advanced TPN in action recognition. © 2020 IEEE.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Time-Aware Traffic Scheduling with Virtual Queues in Time-Sensitive networking

Time-Aware Traffic Scheduling with Virtual Queues in Time-Se...

引用

IFIP/IEEE International Symposium on Integrated network Management

作者： Junli Xue Guochu Shou Yaqiong Liu Yihong Hu Zhigang Guo Beijing Key Laboratory of Network System Architecture and Convergence School of Information and Communication Engineering Beijing University of Posts and Telecommunications Beijing China

Time-Sensitive networking is a promising technology that supports deterministic services with bounded delay, low packet delay variation, and low packet loss. Time-Sensitive networking consists of several components, in which traffic scheduling plays an important role to provide the delay guarantee. The traffic scheduling on the level of flow is challenging for dense traffic demands. In this paper, we propose a time-aware traffic scheduling scheme based on network virtualization to schedule a large number of flows. In the scheme, the physical queues of a network node are abstracted as virtual queues, which isolate each flow with a time offset parameter. A scheduling algorithm is designed into two stages of virtual queues mapping and schedule computation. The simulation results in a use case of industrial automation show the effectiveness of the proposed scheme.

关键词： Schedules Job shop scheduling Automation Scheduling algorithms Simulation Packet loss Delays

来源：评论

学校读者我要写书评

暂无评论

Exploiting visual semantic reasoning for video-text retrieval 29

Exploiting visual semantic reasoning for video-text retrieva...

引用

29th International Joint Conference on Artificial Intelligence, IJCAI 2020

作者： Feng, Zerun Zeng, Zhimin Guo, Caili Li, Zheng Beijing Key Laboratory of Network System Architecture and Convergence School of Information and Communication Engineering Beijing University of Posts and Telecommunications Beijing China Beijing Laboratory of Advanced Information Networks Beijing China

ISBN: (纸本)9780999241165

Video retrieval is a challenging research topic bridging the vision and language areas and has attracted broad attention in recent years. Previous works have been devoted to representing videos by directly encoding from frame-level features. In fact, videos consist of various and abundant semantic relations to which existing methods pay less attention. To address this issue, we propose a Visual Semantic Enhanced Reasoning network (ViSERN) to exploit reasoning between frame regions. Specifically, we consider frame regions as vertices and construct a fully-connected semantic correlation graph. Then, we perform reasoning by novel random walk rule-based graph convolutional networks to generate region features involved with semantic relations. With the benefit of reasoning, semantic interactions between regions are considered, while the impact of redundancy is suppressed. Finally, the region features are aggregated to form frame-level features for further encoding to measure video-text similarity. Extensive experiments on two public benchmark datasets validate the effectiveness of our method by achieving state-ofthe-art performance due to the powerful semantic reasoning. © 2020 Inst. Sci. inf., Univ. Defence in Belgrade. All rights reserved.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

Effective Fusion of 3DCNN and Convolutional GRU for Gesture Recognition 6

Effective Fusion of 3DCNN and Convolutional GRU for Gesture ...

引用

6th IEEE International Conference on Computer and Communications, ICCC 2020

作者： Yangbiao, Luo Muqing, Wu Weiyao, Xu Beijing University of Posts and Telecommunications School of Information and Communication Engineering Beijing Laboratory of Advanced Information Networks Beijing Key Laboratory of Network System Architecture and Convergence Acknowledgment Beijing China

ISBN: (纸本)9781728186351

Gesture recognition is a challenging topic in the field of computer vision. This paper combines the advantages of 3DCNN and ConvRNN in spatial and temporal feature extraction respectively and proposes a network structure of 'Res3DDilated+ConvGRU+2DCNN'. After that, a variant of ConvGRU is proposed, which further exerts the advantages of ConvGRU in temporal feature extraction. The experiment compares the effect of combining dilated convolution in the temporal dimension with 3D Resnet, and the effectiveness of the ConvGRU variant in spatiotemporal information fusion. In the 3D dilated convolutional network, we designed a suitable convolution kernel structure and step size so that subsequent modules can be reused directly. The modular structure design can be easily replaced. The comparison on the IsoGD dataset shows our network effectively integrates spatial and long-short terms features in gesture recognition videos. The variant of ConvGRU also enables each module to play its own advantages in the framework. © 2020 IEEE.

关键词： Convolution

来源：评论

学校读者我要写书评

暂无评论

Study on Pedestrian Detection Based on an Improved YOLOv4 Algorithm 6

Study on Pedestrian Detection Based on an Improved YOLOv4 Al...

引用

6th IEEE International Conference on Computer and Communications, ICCC 2020

作者： Wen, Boyuan Wu, Muqing Beijing University of Posts and Telecommunications School of Information and Communication Engineering Beijing Laboratory of Advanced Information Networks Beijing Key Laboratory of Network System Architecture and Convergence Acknowledgment Beijing China

ISBN: (数字)9781728186351

ISBN: (纸本)9781728186351

Pedestrian detection, which is widely used in automatic driving and pedestrian analysis, has always been a hot research topic in the fields of artificial intelligence and computer vision. With the development of deep learning, pedestrian detectors are becoming more accurate and faster. However, most of them can't strike a balance between accuracy and speed well. Therefore, in this study, we propose a pedestrian detection model based on an improved YOLOv4 algorithm which concerns both detection accuracy and efficiency. The detection model combines a new type of SPP (Spatial Pyramid Pooling) network and K-means clustering algorithm with YOLOv4 model for easier feature extraction. Furthermore, Mish activation function is applied in the neck of the detection model, replacing Leaky ReLU activation function to improve the detection performance. Our pedestrian detector achieves excellent results on the Caltech pedestrian dataset: 84.7% AP at a real-time speed of 36.4 FPS on Titan XP. © 2020 IEEE.

关键词： K-means clustering

来源：评论

学校读者我要写书评

暂无评论

Joint Latency and Energy Consumption Optimization with Deep Reinforcement Learning for Proximity Detection in Road networks

Joint Latency and Energy Consumption Optimization with Deep ...

引用

International Conference on Computer and Communications (ICCC)

作者： Tongyu Zhao Yaqiong Liu Guochu Shou Yihong Hu Beijing Key Laboratory of Network System Architecture and Convergence School of Information and Communication Engineering Beijing University of Posts and Telecommunications Beijing China

ISBN: (纸本)9781665409513

Artificial intelligence and the automobile business have grown significantly in recent years, and autonomous driving has progressively been the industry's focus. The problem of proximity detection in road networks refers to determining if two moving objects are close to one other in real time. However, in the real world, mobile devices' battery life and computing capabilities are restricted, resulting in excessive latency and energy usage. As a result, determining the proximity relationship between mobile users with low latency and energy usage is a difficult task. We formalize the joint latency and energy consumption optimization problem for proximity detection in road networks into a constrained optimization problem (COP) and solve it using a deep Q network (DQN). The simulation results show that DQN can ensure the desired sum cost under a variety of important factors, effectively reducing latency and energy usage.

关键词： Energy consumption Roads Computational modeling Simulation Reinforcement learning Real-time systems Mobile handsets

来源：评论

学校读者我要写书评

暂无评论

An Offloading Scheme Leveraging on Neighboring Node Resources for Edge Computing over Fiber-Wireless (FiWi) Access networks

引用

China Communications 2019年第11期16卷 107-119页

作者： Wei Chang Yihong Hu Guochu Shou Yaqiong Liu Zhigang Guo Beijing Key Laboratory of Network System Architecture and Convergence School of Information and Communication Engineering Beijing University of Posts and Telecommunications

The computation resources at a single node in Edge Computing(EC)are commonly limited,which cannot execute large scale computation *** face the challenge,an Offloading scheme leveraging on NEighboring node Resources(ONER)for EC over Fiber-Wireless(FiWi)access networks is proposed in this *** the ONER scheme,the FiWi network connects edge computing nodes with fiber and converges wireless and fiber connections seamlessly,so that it can support the offloading transmission with low delay and wide *** on the ONER scheme supported by FiWi networks,computation tasks can be offloaded to edge computing nodes in a wider range of area without increasing wireless hops(e.g.,just one wireless hop),which achieves low ***,an efficient Computation Resource Scheduling(CRS)algorithm based on the ONER scheme is also proposed to make offloading *** results show that more offloading requests can be satisfied and the average completion time of computation tasks decreases significantly with the ONER scheme and the CRS ***,the ONER scheme and the CRS algorithm can schedule computation resources at neighboring edge computing nodes for offloading to meet the challenge of large scale computation tasks.

关键词： edge computing offloading Fiber-wireless access networks delay

来源：评论

学校读者我要写书评

暂无评论

An SDN Oriented Loop Detection Mechanism Based on TTL Statistics

引用

China Communications 2020年第6期17卷 1-12页

作者： Tao Yu Longfei Yu Diyue Chen Hongyan Cui Jilong Wang Institute of Network Science and Cyberspace Tsinghua UniversityBeijing 100084China Beijing Advanced Information Network Key Laboratory School of Information and Communication EngineeringBeijing University of Posts and TelecommunicationsBeijing 100876China Network System Architecture and Integration Beijing Key Laboratory Beijing 100876China

The rise of the network has injected new impetus into the development of traditional *** to the complexity of the network itself and its network programmability,there is a risk of routing loops occurring in the SDN *** paper proposes a loop detection *** to the Time To Live(TTL)value of the loop packet,there is approximately periodicity in the same *** use sFlow to count the number of packets corresponding to each TTL value of a switch in the loop over a period of time,and perform discrete Fourier transform on the obtained finite-length sequence to observe its frequency domain performance and determine whether there are periodic *** doing so,it is determined whether there is a routing loop,and the purpose of passively detecting the routing loop is *** to existing algorithms,it has advantages in real-time,scalability and false positive *** experimental results show that the routing loop detection algorithm based on TTL statistics in this paper still maintains high judgment accuracy under the scenarios of lower stream sampling rate and smaller detection period.

关键词： SDN Loop Detection TTL sFlow

来源：评论

学校读者我要写书评

暂无评论

Wireless Channel Modeling, Simulation and Analysis for Electronic Registration Identification of Vehicles 3

Wireless Channel Modeling, Simulation and Analysis for Elect...

引用

3rd IEEE International Conference on Electronic Information and Communication Technology, ICEICT 2020

作者： Cai, Tingxiao Fan, Wenmin Feng, Jing Hong, Weijun China Electronics Standardization Institute Beijing China Beijing Laboratory of Advanced Information Network Beijing University of Posts and Telecommunications Beijing China Beijing Key Laboratory of Network System Architecture and Convergence Beijing University of Posts and Telecommunications Beijing China

ISBN: (纸本)9781728190457

Electronic registration identification (ERI) of the motor vehicle is an important application of passive radio frequency identification (RFID) in the field of intelligent transportation which has been put into use in some countries. ERI is in the pilot research stage now in China, but the release of the standard is accelerating its promotion. The performance evaluation of car-related RFID products is often accomplished by test and trial method currently. The test results are seriously affected by irrelevant environmental factors, making it impossible to carry out scientific repeatable experiments. In this paper, the communication channel model of ERI system in the highway traffic scene is modeled based on the ray tracing method. In order to accurately establish the model, we conducted a simulation analysis on the influence of factors such as antenna pattern, polarization mode, ground material, and environmental noise on the received signal power and tag reading distance. Files generated based on the model in this article can be loaded onto the channel simulator, so as to build an ERI test system with reproducible test results which has important application value. © 2020 IEEE.

关键词： Ray tracing

来源：评论

学校读者我要写书评

暂无评论

Channel Model for RFID based Electronic Registration Identification of the motor vehicle in Practical Open Transportation Scenario 3

Channel Model for RFID based Electronic Registration Identif...

引用

3rd IEEE International Conference on Electronic Information and Communication Technology, ICEICT 2020

作者： Feng, Jing Fan, Wenmin Hong, Weijun Cai, Tingxiao Shi, Chunteng Zhou, Jitianbai China Electronics Standardization Institute Beijing China Beijing Laboratory of Advanced Information Network Beijing University of Posts and Telecommunications Beijing China Beijing Key Laboratory of Network System Architecture and Convergence Beijing University of Posts and Telecommunications Beijing China

ISBN: (纸本)9781728190457

Current researches on communication channel of electronic registration identification of the motor vehicle (ERI) are mostly based on the single lane two channel model, which is not consistent with the multi-lane multi-car scene in the practical application. In this paper, a channel model combined with deterministic and stochastic component is firstly proposed for ERI. The theoretical formula of the model is derived and simulated. The influence of Doppler shift on each transmission path in the channel model and the effect of adjacent lane reflection path on channel fading is analyzed in detail. The work of this paper is with great significance for evaluation of ERI system in practical scenario. © 2020 IEEE.

关键词： Stochastic systems

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：