The motivation for this study is that the quality of deep fakes is constantly improving,which leads to the need to develop new methods for their detection.The proposed Customized Convolutional Neural Network method in...The motivation for this study is that the quality of deep fakes is constantly improving,which leads to the need to develop new methods for their detection.The proposed Customized Convolutional Neural Network method involves extracting structured data from video frames using facial landmark detection,which is then used as input to the CNN.The customized Convolutional Neural Network method is the date augmented-based CNN model to generate‘fake data’or‘fake images’.This study was carried out using Python and its libraries.We used 242 films from the dataset gathered by the Deep Fake Detection Challenge,of which 199 were made up and the remaining 53 were real.Ten seconds were allotted for each video.There were 318 videos used in all,199 of which were fake and 119 of which were real.Our proposedmethod achieved a testing accuracy of 91.47%,loss of 0.342,and AUC score of 0.92,outperforming two alternative approaches,CNN and MLP-CNN.Furthermore,our method succeeded in greater accuracy than contemporary models such as XceptionNet,Meso-4,EfficientNet-BO,MesoInception-4,VGG-16,and DST-Net.The novelty of this investigation is the development of a new Convolutional Neural Network(CNN)learning model that can accurately detect deep fake face photos.展开更多
Neural Networks (NN) are the functional unit of Deep Learning and are known to mimic the behavior of the human brain to solve complex data-driven problems. Whenever we train our own neural networks, we need to take ca...Neural Networks (NN) are the functional unit of Deep Learning and are known to mimic the behavior of the human brain to solve complex data-driven problems. Whenever we train our own neural networks, we need to take care of something called the generalization of the neural network. The performance of Artificial Neural Networks (ANN) mostly depends upon its generalization capability. In this paper, we propose an innovative approach to enhance the generalization capability of artificial neural networks (ANN) using structural redundancy. A novel perspective on handling input data prototypes and their impact on the development of generalization, which could improve to ANN architectures accuracy and reliability is described.展开更多
Regular exercise is a crucial aspect of daily life, as it enables individuals to stay physically active, lowers thelikelihood of developing illnesses, and enhances life expectancy. The recognition of workout actions i...Regular exercise is a crucial aspect of daily life, as it enables individuals to stay physically active, lowers thelikelihood of developing illnesses, and enhances life expectancy. The recognition of workout actions in videostreams holds significant importance in computer vision research, as it aims to enhance exercise adherence, enableinstant recognition, advance fitness tracking technologies, and optimize fitness routines. However, existing actiondatasets often lack diversity and specificity for workout actions, hindering the development of accurate recognitionmodels. To address this gap, the Workout Action Video dataset (WAVd) has been introduced as a significantcontribution. WAVd comprises a diverse collection of labeled workout action videos, meticulously curated toencompass various exercises performed by numerous individuals in different settings. This research proposes aninnovative framework based on the Attention driven Residual Deep Convolutional-Gated Recurrent Unit (ResDCGRU)network for workout action recognition in video streams. Unlike image-based action recognition, videoscontain spatio-temporal information, making the task more complex and challenging. While substantial progresshas been made in this area, challenges persist in detecting subtle and complex actions, handling occlusions,and managing the computational demands of deep learning approaches. The proposed ResDC-GRU Attentionmodel demonstrated exceptional classification performance with 95.81% accuracy in classifying workout actionvideos and also outperformed various state-of-the-art models. The method also yielded 81.6%, 97.2%, 95.6%, and93.2% accuracy on established benchmark datasets, namely HMDB51, Youtube Actions, UCF50, and UCF101,respectively, showcasing its superiority and robustness in action recognition. The findings suggest practicalimplications in real-world scenarios where precise video action recognition is paramount, addressing the persistingchallenges in the field. TheWAVd dataset serves as a catalyst for the development ofmore robust and effective fitnesstracking systems and ultimately promotes healthier lifestyles through improved exercise monitoring and analysis.展开更多
Video summarization aims to select key frames or key shots to create summaries for fast retrieval,compression,and efficient browsing of videos.Graph neural networks efficiently capture information about graph nodes an...Video summarization aims to select key frames or key shots to create summaries for fast retrieval,compression,and efficient browsing of videos.Graph neural networks efficiently capture information about graph nodes and their neighbors,but ignore the dynamic dependencies between nodes.To address this challenge,we propose an innovative Adaptive Graph Convolutional Adjacency Matrix Network(TAMGCN),leveraging the attention mechanism to dynamically adjust dependencies between graph nodes.Specifically,we first segment shots and extract features of each frame,then compute the representative features of each shot.Subsequently,we utilize the attention mechanism to dynamically adjust the adjacency matrix of the graph convolutional network to better capture the dynamic dependencies between graph nodes.Finally,we fuse temporal features extracted by Bi-directional Long Short-Term Memory network with structural features extracted by the graph convolutional network to generate high-quality summaries.Extensive experiments are conducted on two benchmark datasets,TVSum and SumMe,yielding F1-scores of 60.8%and 53.2%,respectively.Experimental results demonstrate that our method outperforms most state-of-the-art video summarization techniques.展开更多
In recent years,real-time video streaming has grown in popularity.The growing popularity of the Internet of Things(IoT)and other wireless heterogeneous networks mandates that network resources be carefully apportioned...In recent years,real-time video streaming has grown in popularity.The growing popularity of the Internet of Things(IoT)and other wireless heterogeneous networks mandates that network resources be carefully apportioned among versatile users in order to achieve the best Quality of Experience(QoE)and performance objectives.Most researchers focused on Forward Error Correction(FEC)techniques when attempting to strike a balance between QoE and performance.However,as network capacity increases,the performance degrades,impacting the live visual experience.Recently,Deep Learning(DL)algorithms have been successfully integrated with FEC to stream videos across multiple heterogeneous networks.But these algorithms need to be changed to make the experience better without sacrificing packet loss and delay time.To address the previous challenge,this paper proposes a novel intelligent algorithm that streams video in multi-home heterogeneous networks based on network-centric characteristics.The proposed framework contains modules such as Intelligent Content Extraction Module(ICEM),Channel Status Monitor(CSM),and Adaptive FEC(AFEC).This framework adopts the Cognitive Learning-based Scheduling(CLS)Module,which works on the deep Reinforced Gated Recurrent Networks(RGRN)principle and embeds them along with the FEC to achieve better performances.The complete framework was developed using the Objective Modular Network Testbed in C++(OMNET++),Internet networking(INET),and Python 3.10,with Keras as the front end and Tensorflow 2.10 as the back end.With extensive experimentation,the proposed model outperforms the other existing intelligentmodels in terms of improving the QoE,minimizing the End-to-End Delay(EED),and maintaining the highest accuracy(98%)and a lower Root Mean Square Error(RMSE)value of 0.001.展开更多
On the basis of measuring the regional high-quality development in China from 2011 to 2020,this study uses gravity model to build spatial correlation network,and uses social network analysis method to analyze the stru...On the basis of measuring the regional high-quality development in China from 2011 to 2020,this study uses gravity model to build spatial correlation network,and uses social network analysis method to analyze the structural characteristics and influencing factors of correlation network.The results are shown as follows.First,from 2011 to 2020,the level of regional high-quality development in China is rising gradually,and the discrete characteristics between regions are gradually obvious,showing a step-like distribution structure decreasing from east to west.Second,the network density of regional high-quality development is generally low and tends to decline,but it has strong stability and correlation strength.Third,the spatial correlation network has an obvious core-edge structure.Shanghai is always at the center of the network and plays a significant intermediary role,while Qinghai and Xinjiang are always at the edge of the network.Fourth,the regional high-quality development association network can be divided into four major sectors:main benefit,net benefit,net spillover,and broker,showing the spatial correlation characteristics of inter-plate contact and intra-plate agglomeration.Fifth,the level of economic development,the level of urbanization and geographical proximity have a significant impact on the formation of regional high-quality development correlation network.展开更多
Recently,the combination of video services and 5G networks have been gaining attention in the wireless communication realm.With the brisk advancement in 5G network usage and the massive popularity of threedimensional ...Recently,the combination of video services and 5G networks have been gaining attention in the wireless communication realm.With the brisk advancement in 5G network usage and the massive popularity of threedimensional video streaming,the quality of experience(QoE)of video in 5G systems has been receiving overwhelming significance from both customers and service provider ends.Therefore,effectively categorizing QoE-aware video streaming is imperative for achieving greater client satisfaction.This work makes the following contribution:First,a simulation platform based on NS-3 is introduced to analyze and improve the performance of video services.The simulation is formulated to offer real-time measurements,saving the expensive expenses associated with real-world equipment.Second,A valuable framework for QoE-aware video streaming categorization is introduced in 5G networks based on machine learning(ML)by incorporating the hyperparameter tuning(HPT)principle.It implements an enhanced hyperparameter tuning(EHPT)ensemble and decision tree(DT)classifier for video streaming categorization.The performance of the ML approach is assessed by considering precision,accuracy,recall,and computation time metrics for manifesting the superiority of these classifiers regarding video streaming categorization.This paper demonstrates that our ML classifiers achieve QoE prediction accuracy of 92.59%for(EHPT)ensemble and 87.037%for decision tree(DT)classifiers.展开更多
At present,the interpretation of regional economic development(RED)has changed from a simple evaluation of economic growth to a focus on economic growth and the optimization of economic structure,the improvement of ec...At present,the interpretation of regional economic development(RED)has changed from a simple evaluation of economic growth to a focus on economic growth and the optimization of economic structure,the improvement of economic relations,and the change of institutional innovation.This article uses the RED trend as the research object and constructs the RED index to conduct the theoretical analysis.Then this paper uses the attention mechanism based on digital twins and the time series network model to verify the actual data.Finally,the regional economy is predicted according to the theoretical model.The specific research work mainly includes the following aspects:1)This paper introduced the development status of research on time series networks and economic forecasting at home and abroad.2)This paper introduces the basic principles and structures of long and short-term memory(LSTM)and convolutional neural network(CNN),constructs an improved CNN-LSTM model combined with the attention mechanism,and then constructs a regional economic prediction index system.3)The best parameters of the model are selected through experiments,and the trained model is used for simulation experiment prediction.The results show that the CNN-LSTM model based on the attentionmechanism proposed in this paper has high accuracy in predicting regional economies.展开更多
The uncertainty of distributed generation energy has dramatically challenged the coordinated development of distribution networks at all levels.This paper focuses on the multi-time-scale regulation model of distribute...The uncertainty of distributed generation energy has dramatically challenged the coordinated development of distribution networks at all levels.This paper focuses on the multi-time-scale regulation model of distributed generation energy under normal conditions.The simulation results of the example verify the self-optimization characteristics and the effectiveness of real-time dispatching of the distribution network control technology at all levels under multiple time scales.展开更多
Currently,the video captioning models based on an encoder-decoder mainly rely on a single video input source.The contents of video captioning are limited since few studies employed external corpus information to guide...Currently,the video captioning models based on an encoder-decoder mainly rely on a single video input source.The contents of video captioning are limited since few studies employed external corpus information to guide the generation of video captioning,which is not conducive to the accurate descrip-tion and understanding of video content.To address this issue,a novel video captioning method guided by a sentence retrieval generation network(ED-SRG)is proposed in this paper.First,a ResNeXt network model,an efficient convolutional network for online video understanding(ECO)model,and a long short-term memory(LSTM)network model are integrated to construct an encoder-decoder,which is utilized to extract the 2D features,3D features,and object features of video data respectively.These features are decoded to generate textual sentences that conform to video content for sentence retrieval.Then,a sentence-transformer network model is employed to retrieve different sentences in an external corpus that are semantically similar to the above textual sentences.The candidate sentences are screened out through similarity measurement.Finally,a novel GPT-2 network model is constructed based on GPT-2 network structure.The model introduces a designed random selector to randomly select predicted words with a high probability in the corpus,which is used to guide and generate textual sentences that are more in line with human natural language expressions.The proposed method in this paper is compared with several existing works by experiments.The results show that the indicators BLEU-4,CIDEr,ROUGE_L,and METEOR are improved by 3.1%,1.3%,0.3%,and 1.5%on a public dataset MSVD and 1.3%,0.5%,0.2%,1.9%on a public dataset MSR-VTT respectively.It can be seen that the proposed method in this paper can generate video captioning with richer semantics than several state-of-the-art approaches.展开更多
For intelligent surveillance videos,anomaly detection is extremely important.Deep learning algorithms have been popular for evaluating realtime surveillance recordings,like traffic accidents,and criminal or unlawful i...For intelligent surveillance videos,anomaly detection is extremely important.Deep learning algorithms have been popular for evaluating realtime surveillance recordings,like traffic accidents,and criminal or unlawful incidents such as suicide attempts.Nevertheless,Deep learning methods for classification,like convolutional neural networks,necessitate a lot of computing power.Quantum computing is a branch of technology that solves abnormal and complex problems using quantum mechanics.As a result,the focus of this research is on developing a hybrid quantum computing model which is based on deep learning.This research develops a Quantum Computing-based Convolutional Neural Network(QC-CNN)to extract features and classify anomalies from surveillance footage.A Quantum-based Circuit,such as the real amplitude circuit,is utilized to improve the performance of the model.As far as my research,this is the first work to employ quantum deep learning techniques to classify anomalous events in video surveillance applications.There are 13 anomalies classified from the UCF-crime dataset.Based on experimental results,the proposed model is capable of efficiently classifying data concerning confusion matrix,Receiver Operating Characteristic(ROC),accuracy,Area Under Curve(AUC),precision,recall as well as F1-score.The proposed QC-CNN has attained the best accuracy of 95.65 percent which is 5.37%greater when compared to other existing models.To measure the efficiency of the proposed work,QC-CNN is also evaluated with classical and quantum models.展开更多
The Global Energy Interconnection is an important strategic approach used to achieve efficient worldwide energy allocation.The idea of developing integrated power,information,and transportation networks provides incre...The Global Energy Interconnection is an important strategic approach used to achieve efficient worldwide energy allocation.The idea of developing integrated power,information,and transportation networks provides increased power interconnection functionality and meaning,helps condense forces,and accelerates the integration of global infrastructure.Correspondingly,it is envisaged that it will become the trend of industrial technological development in the future.In consideration of the current trend of integrated development,this study evaluates a possible plan of coordinated development of fiber-optic and power networks in the Pan-Arctic region.Firstly,the backbone network architecture of Global Energy Interconnection is introduced and the importance of the Arctic energy backbone network is confirmed.The energy consumption and developmental trend of global data centers are then analyzed.Subsequently,the global network traffic is predicted and analyzed by means of a polynomial regression model.Finally,in combination with the current construction of fiber-optic networks in the Pan-Arctic region,the advantages of the integration of the fiber-optic and power networks in this region are clarified in justification of the decision for the development of a Global Energy Interconnection scheme.展开更多
This white paper explores three popular development methodologies for network softwarization: DevOps, NetOps, and Verification. The paper compares and contrasts the strengths and weaknesses of each approach and provid...This white paper explores three popular development methodologies for network softwarization: DevOps, NetOps, and Verification. The paper compares and contrasts the strengths and weaknesses of each approach and provides recommendations for organizations looking to adopt network softwarization.展开更多
Previous studies on brain functional connectivity networks in children have mainly focused on changes in function in specific brain regions, as opposed to whole brain connectivity in healthy children. By analyzing the...Previous studies on brain functional connectivity networks in children have mainly focused on changes in function in specific brain regions, as opposed to whole brain connectivity in healthy children. By analyzing the independent components of activation and network connectivity between brain regions, we examined brain activity status and development trends in children aged 3 and 5 years. These data could provide a reference for brain function rehabilitation in children with illness or abnormal function. We acquired functional magnetic resonance images from 15 3-year-old children and 15 5-year-old children under natural sleep cond让ions. The participants were recruited from five kindergartens in the Nanshan District of Shenzhen City, China. The parents of the participants signed an informed consent form with the premise that they had been fully informed regarding the experimental protocol. We used masked independent component analysis and BrainNet Viewer software to explore the independent components of the brain and correlation connections between brain regions. We identified seven independent components in the two groups of children, including the executive control network, the dorsal attention network, the default mode network, the left frontoparietal network, the right frontoparietal network, the salience network, and the motor network. In the default mode network, the posterior cingulate cortex, medial frontal gyrus, and inferior parietal lobule were activated in both 3- and 5-year-old children, supporting the "three-brain region theory” of the default mode network. In the frontoparietal network, the frontal and parietal gyri were activated in the two groups of children, and functional connectivity was strengthened in 5-year-olds compared with 3-year-olds, although the nodes and network connections were not yet mature. The high-correlation network connections in the default mode networks and dorsal attention networks had been significantly strengthened in 5-year-olds vs. 3-year-olds. Further, the salience network in the 3-year-old children included an activated insula/inferior frontal gyrus-anterior cingulate cortex network circu让 and an activated thalamus-parahippocampal-posterior cingulate cortex-subcortical regions network circuit. By the age of 5 years, no des and high-correlation network connections (edges) were reduced in the salience network. Overall, activation of the dorsal attention network, default mode network, left frontoparietal network, and right frontoparietal network increased (the volume of activation increased, the signals strengthened, and the high-correlation connections increased and strengthened) in 5-year-olds compared with 3-year-olds, but activation in some brain nodes weakened or disappeared in the salience network, and the network connections (edges) were reduced. Between the ages of 3 and 5 years, we observed a tendency for function in some brain regions to be strengthened and for the generalization of activation to be reduced, indicating that specialization begins to develop at this time. The study protocol was approved by the local ethics committee of the Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences in China with approval No. SIAT-IRB- 131115-H0075 on November 15, 2013.展开更多
The increasing popularity of smart mobile devices and the rise of online services has increased the requirements for efficient dissemination of social video contents. In this paper,we study the problem of distributing...The increasing popularity of smart mobile devices and the rise of online services has increased the requirements for efficient dissemination of social video contents. In this paper,we study the problem of distributing video from cloud server to users in partially connected cooperative D2 D network using network coding. In such a scenario, the transmission conflicts occur from simultaneous transmissions of multiple devices, and the scheduling decision should be made not only on the encoded packets but also on the set of transmitting devices. We analyze the lower bound and give an integer linear formulation of the joint optimization problem over the set of transmitting devices and the packet combinations.We also propose a heuristic solution for this setup using a conflict graph and local graph at every device. Simulation results show that our coding scheme significantly reduces the number of transmission slots, which will increase the efficiency of video delivery.展开更多
Most of previous video recording devices in mobile vehicles commonly store captured video contents locally. With the rapid development of 4G/Wi Fi networks, there emerges a new trend to equip video recording devices w...Most of previous video recording devices in mobile vehicles commonly store captured video contents locally. With the rapid development of 4G/Wi Fi networks, there emerges a new trend to equip video recording devices with wireless interfaces to enable video uploading to the cloud for video playback in a later time point. In this paper, we propose a QoE-aware mobile cloud video recording scheme in the roadside vehicular networks, which can adaptively select the proper wireless interface and video bitrate for video uploading to the cloud. To maximize the total utility, we need to design a control strategy to carefully balance the transmission cost and the achieved QoE for users. To this purpose, we investigate the tradeoff between cost incurred by uploading through cellular networks and the achieved QoE of users. We apply the optimization framework to solve the formulated problem and design an online scheduling algorithm. We also conduct extensive trace-driven simulations and our results show that our algorithm achieves a good balance between the transmission cost and user QoE.展开更多
Scalable video coding(SVC)has been widely used in video-on-demand(VOD)service,to efficiently satisfy users’different video quality requirements and dynamically adjust video stream to timevariant wireless channels.Und...Scalable video coding(SVC)has been widely used in video-on-demand(VOD)service,to efficiently satisfy users’different video quality requirements and dynamically adjust video stream to timevariant wireless channels.Under the 5G network structure,we consider a cooperative caching scheme inside each cluster with SVC to economically utilize the limited caching storage.A novel multi-agent deep reinforcement learning(MADRL)framework is proposed to jointly optimize the video access delay and users’satisfaction,where an aggregation node is introduced helping individual agents to achieve global observations and overall system rewards.Moreover,to cope with the large action space caused by the large number of videos and users,a dimension decomposition method is embedded into the neural network in each agent,which greatly reduce the computational complexity and memory cost of the reinforcement learning.Experimental results show that:1)the proposed value-decomposed dimensional network(VDDN)algorithm achieves an obvious performance gain versus the traditional MADRL;2)the proposed VDDN algorithm can handle an extremely large action space and quickly converge with a low computational complexity.展开更多
To cope with the rapid growth of mobile video, video providers have leveraged cloud technologies to deploy their mobile video service system for more cost-effective and scalable performance. The emergence of Software-...To cope with the rapid growth of mobile video, video providers have leveraged cloud technologies to deploy their mobile video service system for more cost-effective and scalable performance. The emergence of Software-Defined Networking(SDN) provides a promising solution to manage the underlying network. In this paper, we introduce an SDN-enabled cloud mobile video distribution architecture and propose a joint video placement, request dispatching and traffic management mechanism to improve user experience and reduce the system operational cost. We use a utility function to capture the two aspects of user experience: the level of satisfaction and average latency, and formulate the joint optimization problem as a mixed integer programming problem. We develop an optimal algorithm based on dual decomposition and prove its optimality. We conduct simulations to evaluate the performance of our algorithm and the results show that our strategy can effectively cut down the total cost and guarantee user experience.展开更多
To evaluate the video quality, we tested sample videos delivered using HTTP adaptive streaming (HAS) in LTE network. In order to establish a correlation between radio access network (RAN) performance and quality o...To evaluate the video quality, we tested sample videos delivered using HTTP adaptive streaming (HAS) in LTE network. In order to establish a correlation between radio access network (RAN) performance and quality of experience ( QoE), we set up a testbed under different radio im- pairment conditions with three parameters: signal to interference and noise ratio ( SINR), an amount of available network resource and a round trip latency. End users graded each video in a mobile equipment with their QoE Mearnwhile, we used a nonlinear model to simulate the comprehensive pre- dicted mean opinion score (pMOS). Our results show that the nonlinear model can predict the enduser' s feedback. The pearson correlation coefficient (PCC) of the model is larger than 0. 9. This demonstrate that the output of the model has a high correlation with the end users' ratings and can reflect the QoE accurately. The method we developed will help mobile network operators evaluate the RAN performance of its QoE. It can also be used for HAS service to optimize LTE network and improve its QoE.展开更多
Today the cycle time of the product develop is requ ir ed to be shortened. At the same time the requirement of the customers becomes mo re and more diverse and complex. The capability of the develop unit is limited b ...Today the cycle time of the product develop is requ ir ed to be shortened. At the same time the requirement of the customers becomes mo re and more diverse and complex. The capability of the develop unit is limited b ecause of the existence of heterogeneous systems and distributed environments. I n this paper, we bring forward a new approach to solve the problem in product de velopment process. We also settle part key technologies in it. A great deal of information from all kinds of sources in the distributed develop ment process is interweaved. The solution to organize the workflow and manage th e information in the process is called for anxiously. We use a new approach that is asynchronous and synchronous coupling product development approach based on the network. The approach extends the develop process from the time axis. Then t he activities in the process are organized from the asynchronous and synchronous aspects. The state of every activity projects at the ASN (active semantic netwo rk). The ASN includes decision system, intelligent agent, user interface and net work. The ASN decides the types and states of the activities and deals with the couple relationship among them. The knowledge stored in ASN is open to all users through the relative interfaces. Every specialist keeps contact with their user s relying on collaborative platform implements CSCW (computer support collaborat ive work) that integrated product/process design and development. The lack of gl obal communication in product development process can be prevented in the most d egree. The key technologies that exist in the asynchronous and synchronous coupling pro duct develop approach include: integrated development structure, orderly organiz ation of information, transparent management of process, agile transfer of infor mation and rapid prototype. The development process can be completed quickly by these technologies. The technologies involve wide content. In this paper, we dis cuss some key technologies. We validate the approach by the projectrapid response manufacturing a pplication in the distributed environment. The expensive device, high technology and low using lead to RE (Rapid engineering) and RP (Rapid prototype) service a pplication by the network. RE and RP develop rapidly due to the accelerated prod uct development process. RE and RP application service platform is built in the project.展开更多
基金Science and Technology Funds from the Liaoning Education Department(Serial Number:LJKZ0104).
文摘The motivation for this study is that the quality of deep fakes is constantly improving,which leads to the need to develop new methods for their detection.The proposed Customized Convolutional Neural Network method involves extracting structured data from video frames using facial landmark detection,which is then used as input to the CNN.The customized Convolutional Neural Network method is the date augmented-based CNN model to generate‘fake data’or‘fake images’.This study was carried out using Python and its libraries.We used 242 films from the dataset gathered by the Deep Fake Detection Challenge,of which 199 were made up and the remaining 53 were real.Ten seconds were allotted for each video.There were 318 videos used in all,199 of which were fake and 119 of which were real.Our proposedmethod achieved a testing accuracy of 91.47%,loss of 0.342,and AUC score of 0.92,outperforming two alternative approaches,CNN and MLP-CNN.Furthermore,our method succeeded in greater accuracy than contemporary models such as XceptionNet,Meso-4,EfficientNet-BO,MesoInception-4,VGG-16,and DST-Net.The novelty of this investigation is the development of a new Convolutional Neural Network(CNN)learning model that can accurately detect deep fake face photos.
文摘Neural Networks (NN) are the functional unit of Deep Learning and are known to mimic the behavior of the human brain to solve complex data-driven problems. Whenever we train our own neural networks, we need to take care of something called the generalization of the neural network. The performance of Artificial Neural Networks (ANN) mostly depends upon its generalization capability. In this paper, we propose an innovative approach to enhance the generalization capability of artificial neural networks (ANN) using structural redundancy. A novel perspective on handling input data prototypes and their impact on the development of generalization, which could improve to ANN architectures accuracy and reliability is described.
文摘Regular exercise is a crucial aspect of daily life, as it enables individuals to stay physically active, lowers thelikelihood of developing illnesses, and enhances life expectancy. The recognition of workout actions in videostreams holds significant importance in computer vision research, as it aims to enhance exercise adherence, enableinstant recognition, advance fitness tracking technologies, and optimize fitness routines. However, existing actiondatasets often lack diversity and specificity for workout actions, hindering the development of accurate recognitionmodels. To address this gap, the Workout Action Video dataset (WAVd) has been introduced as a significantcontribution. WAVd comprises a diverse collection of labeled workout action videos, meticulously curated toencompass various exercises performed by numerous individuals in different settings. This research proposes aninnovative framework based on the Attention driven Residual Deep Convolutional-Gated Recurrent Unit (ResDCGRU)network for workout action recognition in video streams. Unlike image-based action recognition, videoscontain spatio-temporal information, making the task more complex and challenging. While substantial progresshas been made in this area, challenges persist in detecting subtle and complex actions, handling occlusions,and managing the computational demands of deep learning approaches. The proposed ResDC-GRU Attentionmodel demonstrated exceptional classification performance with 95.81% accuracy in classifying workout actionvideos and also outperformed various state-of-the-art models. The method also yielded 81.6%, 97.2%, 95.6%, and93.2% accuracy on established benchmark datasets, namely HMDB51, Youtube Actions, UCF50, and UCF101,respectively, showcasing its superiority and robustness in action recognition. The findings suggest practicalimplications in real-world scenarios where precise video action recognition is paramount, addressing the persistingchallenges in the field. TheWAVd dataset serves as a catalyst for the development ofmore robust and effective fitnesstracking systems and ultimately promotes healthier lifestyles through improved exercise monitoring and analysis.
基金This work was supported by Natural Science Foundation of Gansu Province under Grant Nos.21JR7RA570,20JR10RA334Basic Research Program of Gansu Province No.22JR11RA106,Gansu University of Political Science and Law Major Scientific Research and Innovation Projects under Grant No.GZF2020XZDA03+1 种基金the Young Doctoral Fund Project of Higher Education Institutions in Gansu Province in 2022 under Grant No.2022QB-123,Gansu Province Higher Education Innovation Fund Project under Grant No.2022A-097the University-Level Research Funding Project under Grant No.GZFXQNLW022 and University-Level Innovative Research Team of Gansu University of Political Science and Law.
文摘Video summarization aims to select key frames or key shots to create summaries for fast retrieval,compression,and efficient browsing of videos.Graph neural networks efficiently capture information about graph nodes and their neighbors,but ignore the dynamic dependencies between nodes.To address this challenge,we propose an innovative Adaptive Graph Convolutional Adjacency Matrix Network(TAMGCN),leveraging the attention mechanism to dynamically adjust dependencies between graph nodes.Specifically,we first segment shots and extract features of each frame,then compute the representative features of each shot.Subsequently,we utilize the attention mechanism to dynamically adjust the adjacency matrix of the graph convolutional network to better capture the dynamic dependencies between graph nodes.Finally,we fuse temporal features extracted by Bi-directional Long Short-Term Memory network with structural features extracted by the graph convolutional network to generate high-quality summaries.Extensive experiments are conducted on two benchmark datasets,TVSum and SumMe,yielding F1-scores of 60.8%and 53.2%,respectively.Experimental results demonstrate that our method outperforms most state-of-the-art video summarization techniques.
文摘In recent years,real-time video streaming has grown in popularity.The growing popularity of the Internet of Things(IoT)and other wireless heterogeneous networks mandates that network resources be carefully apportioned among versatile users in order to achieve the best Quality of Experience(QoE)and performance objectives.Most researchers focused on Forward Error Correction(FEC)techniques when attempting to strike a balance between QoE and performance.However,as network capacity increases,the performance degrades,impacting the live visual experience.Recently,Deep Learning(DL)algorithms have been successfully integrated with FEC to stream videos across multiple heterogeneous networks.But these algorithms need to be changed to make the experience better without sacrificing packet loss and delay time.To address the previous challenge,this paper proposes a novel intelligent algorithm that streams video in multi-home heterogeneous networks based on network-centric characteristics.The proposed framework contains modules such as Intelligent Content Extraction Module(ICEM),Channel Status Monitor(CSM),and Adaptive FEC(AFEC).This framework adopts the Cognitive Learning-based Scheduling(CLS)Module,which works on the deep Reinforced Gated Recurrent Networks(RGRN)principle and embeds them along with the FEC to achieve better performances.The complete framework was developed using the Objective Modular Network Testbed in C++(OMNET++),Internet networking(INET),and Python 3.10,with Keras as the front end and Tensorflow 2.10 as the back end.With extensive experimentation,the proposed model outperforms the other existing intelligentmodels in terms of improving the QoE,minimizing the End-to-End Delay(EED),and maintaining the highest accuracy(98%)and a lower Root Mean Square Error(RMSE)value of 0.001.
文摘On the basis of measuring the regional high-quality development in China from 2011 to 2020,this study uses gravity model to build spatial correlation network,and uses social network analysis method to analyze the structural characteristics and influencing factors of correlation network.The results are shown as follows.First,from 2011 to 2020,the level of regional high-quality development in China is rising gradually,and the discrete characteristics between regions are gradually obvious,showing a step-like distribution structure decreasing from east to west.Second,the network density of regional high-quality development is generally low and tends to decline,but it has strong stability and correlation strength.Third,the spatial correlation network has an obvious core-edge structure.Shanghai is always at the center of the network and plays a significant intermediary role,while Qinghai and Xinjiang are always at the edge of the network.Fourth,the regional high-quality development association network can be divided into four major sectors:main benefit,net benefit,net spillover,and broker,showing the spatial correlation characteristics of inter-plate contact and intra-plate agglomeration.Fifth,the level of economic development,the level of urbanization and geographical proximity have a significant impact on the formation of regional high-quality development correlation network.
文摘Recently,the combination of video services and 5G networks have been gaining attention in the wireless communication realm.With the brisk advancement in 5G network usage and the massive popularity of threedimensional video streaming,the quality of experience(QoE)of video in 5G systems has been receiving overwhelming significance from both customers and service provider ends.Therefore,effectively categorizing QoE-aware video streaming is imperative for achieving greater client satisfaction.This work makes the following contribution:First,a simulation platform based on NS-3 is introduced to analyze and improve the performance of video services.The simulation is formulated to offer real-time measurements,saving the expensive expenses associated with real-world equipment.Second,A valuable framework for QoE-aware video streaming categorization is introduced in 5G networks based on machine learning(ML)by incorporating the hyperparameter tuning(HPT)principle.It implements an enhanced hyperparameter tuning(EHPT)ensemble and decision tree(DT)classifier for video streaming categorization.The performance of the ML approach is assessed by considering precision,accuracy,recall,and computation time metrics for manifesting the superiority of these classifiers regarding video streaming categorization.This paper demonstrates that our ML classifiers achieve QoE prediction accuracy of 92.59%for(EHPT)ensemble and 87.037%for decision tree(DT)classifiers.
文摘At present,the interpretation of regional economic development(RED)has changed from a simple evaluation of economic growth to a focus on economic growth and the optimization of economic structure,the improvement of economic relations,and the change of institutional innovation.This article uses the RED trend as the research object and constructs the RED index to conduct the theoretical analysis.Then this paper uses the attention mechanism based on digital twins and the time series network model to verify the actual data.Finally,the regional economy is predicted according to the theoretical model.The specific research work mainly includes the following aspects:1)This paper introduced the development status of research on time series networks and economic forecasting at home and abroad.2)This paper introduces the basic principles and structures of long and short-term memory(LSTM)and convolutional neural network(CNN),constructs an improved CNN-LSTM model combined with the attention mechanism,and then constructs a regional economic prediction index system.3)The best parameters of the model are selected through experiments,and the trained model is used for simulation experiment prediction.The results show that the CNN-LSTM model based on the attentionmechanism proposed in this paper has high accuracy in predicting regional economies.
文摘The uncertainty of distributed generation energy has dramatically challenged the coordinated development of distribution networks at all levels.This paper focuses on the multi-time-scale regulation model of distributed generation energy under normal conditions.The simulation results of the example verify the self-optimization characteristics and the effectiveness of real-time dispatching of the distribution network control technology at all levels under multiple time scales.
基金supported in part by the National Natural Science Foundation of China under Grants 62273272 and 61873277in part by the Chinese Postdoctoral Science Foundation under Grant 2020M673446+1 种基金in part by the Key Research and Development Program of Shaanxi Province under Grant 2023-YBGY-243in part by the Youth Innovation Team of Shaanxi Universities.
文摘Currently,the video captioning models based on an encoder-decoder mainly rely on a single video input source.The contents of video captioning are limited since few studies employed external corpus information to guide the generation of video captioning,which is not conducive to the accurate descrip-tion and understanding of video content.To address this issue,a novel video captioning method guided by a sentence retrieval generation network(ED-SRG)is proposed in this paper.First,a ResNeXt network model,an efficient convolutional network for online video understanding(ECO)model,and a long short-term memory(LSTM)network model are integrated to construct an encoder-decoder,which is utilized to extract the 2D features,3D features,and object features of video data respectively.These features are decoded to generate textual sentences that conform to video content for sentence retrieval.Then,a sentence-transformer network model is employed to retrieve different sentences in an external corpus that are semantically similar to the above textual sentences.The candidate sentences are screened out through similarity measurement.Finally,a novel GPT-2 network model is constructed based on GPT-2 network structure.The model introduces a designed random selector to randomly select predicted words with a high probability in the corpus,which is used to guide and generate textual sentences that are more in line with human natural language expressions.The proposed method in this paper is compared with several existing works by experiments.The results show that the indicators BLEU-4,CIDEr,ROUGE_L,and METEOR are improved by 3.1%,1.3%,0.3%,and 1.5%on a public dataset MSVD and 1.3%,0.5%,0.2%,1.9%on a public dataset MSR-VTT respectively.It can be seen that the proposed method in this paper can generate video captioning with richer semantics than several state-of-the-art approaches.
文摘For intelligent surveillance videos,anomaly detection is extremely important.Deep learning algorithms have been popular for evaluating realtime surveillance recordings,like traffic accidents,and criminal or unlawful incidents such as suicide attempts.Nevertheless,Deep learning methods for classification,like convolutional neural networks,necessitate a lot of computing power.Quantum computing is a branch of technology that solves abnormal and complex problems using quantum mechanics.As a result,the focus of this research is on developing a hybrid quantum computing model which is based on deep learning.This research develops a Quantum Computing-based Convolutional Neural Network(QC-CNN)to extract features and classify anomalies from surveillance footage.A Quantum-based Circuit,such as the real amplitude circuit,is utilized to improve the performance of the model.As far as my research,this is the first work to employ quantum deep learning techniques to classify anomalous events in video surveillance applications.There are 13 anomalies classified from the UCF-crime dataset.Based on experimental results,the proposed model is capable of efficiently classifying data concerning confusion matrix,Receiver Operating Characteristic(ROC),accuracy,Area Under Curve(AUC),precision,recall as well as F1-score.The proposed QC-CNN has attained the best accuracy of 95.65 percent which is 5.37%greater when compared to other existing models.To measure the efficiency of the proposed work,QC-CNN is also evaluated with classical and quantum models.
基金supported by the Corporation Science and Technology Program of Global Energy Interconnection Group Ltd. (GEIGC-D-[2018]024)by the National Natural Science Foundation of China (61472042, 61772079)
文摘The Global Energy Interconnection is an important strategic approach used to achieve efficient worldwide energy allocation.The idea of developing integrated power,information,and transportation networks provides increased power interconnection functionality and meaning,helps condense forces,and accelerates the integration of global infrastructure.Correspondingly,it is envisaged that it will become the trend of industrial technological development in the future.In consideration of the current trend of integrated development,this study evaluates a possible plan of coordinated development of fiber-optic and power networks in the Pan-Arctic region.Firstly,the backbone network architecture of Global Energy Interconnection is introduced and the importance of the Arctic energy backbone network is confirmed.The energy consumption and developmental trend of global data centers are then analyzed.Subsequently,the global network traffic is predicted and analyzed by means of a polynomial regression model.Finally,in combination with the current construction of fiber-optic networks in the Pan-Arctic region,the advantages of the integration of the fiber-optic and power networks in this region are clarified in justification of the decision for the development of a Global Energy Interconnection scheme.
文摘This white paper explores three popular development methodologies for network softwarization: DevOps, NetOps, and Verification. The paper compares and contrasts the strengths and weaknesses of each approach and provides recommendations for organizations looking to adopt network softwarization.
基金supported by the Natural Science Foundation of Guangdong Province,No.2016A030313180(to FCJ)
文摘Previous studies on brain functional connectivity networks in children have mainly focused on changes in function in specific brain regions, as opposed to whole brain connectivity in healthy children. By analyzing the independent components of activation and network connectivity between brain regions, we examined brain activity status and development trends in children aged 3 and 5 years. These data could provide a reference for brain function rehabilitation in children with illness or abnormal function. We acquired functional magnetic resonance images from 15 3-year-old children and 15 5-year-old children under natural sleep cond让ions. The participants were recruited from five kindergartens in the Nanshan District of Shenzhen City, China. The parents of the participants signed an informed consent form with the premise that they had been fully informed regarding the experimental protocol. We used masked independent component analysis and BrainNet Viewer software to explore the independent components of the brain and correlation connections between brain regions. We identified seven independent components in the two groups of children, including the executive control network, the dorsal attention network, the default mode network, the left frontoparietal network, the right frontoparietal network, the salience network, and the motor network. In the default mode network, the posterior cingulate cortex, medial frontal gyrus, and inferior parietal lobule were activated in both 3- and 5-year-old children, supporting the "three-brain region theory” of the default mode network. In the frontoparietal network, the frontal and parietal gyri were activated in the two groups of children, and functional connectivity was strengthened in 5-year-olds compared with 3-year-olds, although the nodes and network connections were not yet mature. The high-correlation network connections in the default mode networks and dorsal attention networks had been significantly strengthened in 5-year-olds vs. 3-year-olds. Further, the salience network in the 3-year-old children included an activated insula/inferior frontal gyrus-anterior cingulate cortex network circu让 and an activated thalamus-parahippocampal-posterior cingulate cortex-subcortical regions network circuit. By the age of 5 years, no des and high-correlation network connections (edges) were reduced in the salience network. Overall, activation of the dorsal attention network, default mode network, left frontoparietal network, and right frontoparietal network increased (the volume of activation increased, the signals strengthened, and the high-correlation connections increased and strengthened) in 5-year-olds compared with 3-year-olds, but activation in some brain nodes weakened or disappeared in the salience network, and the network connections (edges) were reduced. Between the ages of 3 and 5 years, we observed a tendency for function in some brain regions to be strengthened and for the generalization of activation to be reduced, indicating that specialization begins to develop at this time. The study protocol was approved by the local ethics committee of the Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences in China with approval No. SIAT-IRB- 131115-H0075 on November 15, 2013.
基金supported by Fundamental Research Funds for the Central Universities(No.SWU115002,No.XDJK2015C104)
文摘The increasing popularity of smart mobile devices and the rise of online services has increased the requirements for efficient dissemination of social video contents. In this paper,we study the problem of distributing video from cloud server to users in partially connected cooperative D2 D network using network coding. In such a scenario, the transmission conflicts occur from simultaneous transmissions of multiple devices, and the scheduling decision should be made not only on the encoded packets but also on the set of transmitting devices. We analyze the lower bound and give an integer linear formulation of the joint optimization problem over the set of transmitting devices and the packet combinations.We also propose a heuristic solution for this setup using a conflict graph and local graph at every device. Simulation results show that our coding scheme significantly reduces the number of transmission slots, which will increase the efficiency of video delivery.
基金supported in part by the National Science Foundation of China under Grant 61272397,Grant 61572538,Grant 61174152,Grant 61331008in part by the Guangdong Natural Science Funds for Distinguished Young Scholar under Grant S20120011187
文摘Most of previous video recording devices in mobile vehicles commonly store captured video contents locally. With the rapid development of 4G/Wi Fi networks, there emerges a new trend to equip video recording devices with wireless interfaces to enable video uploading to the cloud for video playback in a later time point. In this paper, we propose a QoE-aware mobile cloud video recording scheme in the roadside vehicular networks, which can adaptively select the proper wireless interface and video bitrate for video uploading to the cloud. To maximize the total utility, we need to design a control strategy to carefully balance the transmission cost and the achieved QoE for users. To this purpose, we investigate the tradeoff between cost incurred by uploading through cellular networks and the achieved QoE of users. We apply the optimization framework to solve the formulated problem and design an online scheduling algorithm. We also conduct extensive trace-driven simulations and our results show that our algorithm achieves a good balance between the transmission cost and user QoE.
基金supported by the National Natural Science Foundation of China under Grant No.61801119。
文摘Scalable video coding(SVC)has been widely used in video-on-demand(VOD)service,to efficiently satisfy users’different video quality requirements and dynamically adjust video stream to timevariant wireless channels.Under the 5G network structure,we consider a cooperative caching scheme inside each cluster with SVC to economically utilize the limited caching storage.A novel multi-agent deep reinforcement learning(MADRL)framework is proposed to jointly optimize the video access delay and users’satisfaction,where an aggregation node is introduced helping individual agents to achieve global observations and overall system rewards.Moreover,to cope with the large action space caused by the large number of videos and users,a dimension decomposition method is embedded into the neural network in each agent,which greatly reduce the computational complexity and memory cost of the reinforcement learning.Experimental results show that:1)the proposed value-decomposed dimensional network(VDDN)algorithm achieves an obvious performance gain versus the traditional MADRL;2)the proposed VDDN algorithm can handle an extremely large action space and quickly converge with a low computational complexity.
基金supported by the State Key Program of National Natural Science Foundation of China(Grant No.61233003)National Natural Science Foundation of China(Grant No.61503358)
文摘To cope with the rapid growth of mobile video, video providers have leveraged cloud technologies to deploy their mobile video service system for more cost-effective and scalable performance. The emergence of Software-Defined Networking(SDN) provides a promising solution to manage the underlying network. In this paper, we introduce an SDN-enabled cloud mobile video distribution architecture and propose a joint video placement, request dispatching and traffic management mechanism to improve user experience and reduce the system operational cost. We use a utility function to capture the two aspects of user experience: the level of satisfaction and average latency, and formulate the joint optimization problem as a mixed integer programming problem. We develop an optimal algorithm based on dual decomposition and prove its optimality. We conduct simulations to evaluate the performance of our algorithm and the results show that our strategy can effectively cut down the total cost and guarantee user experience.
基金Supported by China National S&T Major Project(2013ZX03003002-003)Beijing Natural Science Foundation(4152047)111Project of China(B14010)
文摘To evaluate the video quality, we tested sample videos delivered using HTTP adaptive streaming (HAS) in LTE network. In order to establish a correlation between radio access network (RAN) performance and quality of experience ( QoE), we set up a testbed under different radio im- pairment conditions with three parameters: signal to interference and noise ratio ( SINR), an amount of available network resource and a round trip latency. End users graded each video in a mobile equipment with their QoE Mearnwhile, we used a nonlinear model to simulate the comprehensive pre- dicted mean opinion score (pMOS). Our results show that the nonlinear model can predict the enduser' s feedback. The pearson correlation coefficient (PCC) of the model is larger than 0. 9. This demonstrate that the output of the model has a high correlation with the end users' ratings and can reflect the QoE accurately. The method we developed will help mobile network operators evaluate the RAN performance of its QoE. It can also be used for HAS service to optimize LTE network and improve its QoE.
文摘Today the cycle time of the product develop is requ ir ed to be shortened. At the same time the requirement of the customers becomes mo re and more diverse and complex. The capability of the develop unit is limited b ecause of the existence of heterogeneous systems and distributed environments. I n this paper, we bring forward a new approach to solve the problem in product de velopment process. We also settle part key technologies in it. A great deal of information from all kinds of sources in the distributed develop ment process is interweaved. The solution to organize the workflow and manage th e information in the process is called for anxiously. We use a new approach that is asynchronous and synchronous coupling product development approach based on the network. The approach extends the develop process from the time axis. Then t he activities in the process are organized from the asynchronous and synchronous aspects. The state of every activity projects at the ASN (active semantic netwo rk). The ASN includes decision system, intelligent agent, user interface and net work. The ASN decides the types and states of the activities and deals with the couple relationship among them. The knowledge stored in ASN is open to all users through the relative interfaces. Every specialist keeps contact with their user s relying on collaborative platform implements CSCW (computer support collaborat ive work) that integrated product/process design and development. The lack of gl obal communication in product development process can be prevented in the most d egree. The key technologies that exist in the asynchronous and synchronous coupling pro duct develop approach include: integrated development structure, orderly organiz ation of information, transparent management of process, agile transfer of infor mation and rapid prototype. The development process can be completed quickly by these technologies. The technologies involve wide content. In this paper, we dis cuss some key technologies. We validate the approach by the projectrapid response manufacturing a pplication in the distributed environment. The expensive device, high technology and low using lead to RE (Rapid engineering) and RP (Rapid prototype) service a pplication by the network. RE and RP develop rapidly due to the accelerated prod uct development process. RE and RP application service platform is built in the project.