利用改进深度信念网络的人脸表情识别

针对人脸表情识别过程中,误差逆向传播算法(back propagation,BP)对深度信念网络(deep belief network,DBN)微调时容易陷入极值点局部极小和收敛时间过长的问题,提出基于BP算法微调DBN的改进方法.对表情进行多特征提取并降维,利用所提方法对降维后特征进行学习,采用共轭梯度算法解决BP...     

Analysis of high-level dance movements under deep learning and internet of things

The Journal of Supercomputing - The purpose is to improve the design effect of high-level dance movements and help dancers to better master these movements. The body changes with advanced dance...     

Image recognition based on improved convolutional deep belief network model

Multimedia Tools and Applications - Aiming at the homogeneity of convolution kernels in Convolutional Deep Belief Network (CDBN), a cross-entropy-based sparse penalty mechanism suitable for...     

机床物联网现场层数据采集终端的设计与实现

为了提高数控机床的运行管理和维护水平,满足现代制造业的信息化需求,设计了一种基于射频(RFID)和传感器的机床物联网现场层数据采集终端,其包括数据采集单元和数据传输单元。该数据采集终端能够将来自各传感器的数据以工业以太网、WiFi、ZigBee三种方式传输到监控与诊断服务器,为远程监控与诊断提供准确可靠的实时信息。     

Multi-input 1-dimensional deep belief network: action and activity recognition as case study

Multimedia Tools and Applications - This paper develops a new variation of deep belief networks which is evaluated on the basis of supervised classification of human actions and activities. The...     

Applying the blockchain-based deep reinforcement consensus algorithm to the intelligent manufacturing model under internet of things

The Journal of Supercomputing - “Industry 4.0”, namely intelligent manufacturing (IM), includes intelligent production (IP) and smart factory (SF). The study aims to improve the...     

动态深度信念网络模型构建

深度学习是一类新兴的多层神经网络学习算法,因其缓解了传统训练算法的局部最小性,故引起机器学习领域的广泛关注。但是,如何使一个网络模型在选取任意数值的隐藏层节点数时都能够得到一个比较合适的网络结构是目前深度学习界普遍存在的一个开放性问题。文章提出了一种能够动态地学习模型结构的算法——最大判别能力转换法,根据Fisher准则来评估隐藏层每一个节点的判别性能,然后通过动态地选择部分隐层节点来构建最优的模型结构。     

Kinematic and dynamic control model of wheeled mobile robot under internet of things and neural network

The Journal of Supercomputing - This study aims to solve the issues of nonlinearity, non-integrity constraints, under-actuated systems in mobile robots. The wheeled robot is selected as the...     

Intelligent multimedia big data computing for internet of things

Multimedia Tools and Applications -     

Bidirectional self-adaptive resampling in internet of things big data learning

Multimedia Tools and Applications - This paper focuses on the problem of low learning algorithm accuracy caused by serious imbalance of big data in Internet of Things, and proposes a bidirectional...     

物联网条件下的信息安全风险防范及立法建议

随着信息技术的发展,物联网在我国得到了广泛的应用。物联网给人们带来巨大便利的同时,也对信息安全造成了巨大的风险。基于对国内外相关立法的比较,文章提出了物联网时代信息安全保护的立法原则、立法体系与结构、立法内容等构想。同时,提出了强化政府的监管作用、逐步统一规范和标准、加大物联网条件下的密码保护力度、提升用户自身安全意识等建议。     

深度信念网络在故障指示器检测中的应用

电子产品在生产过程中需进行产品检测,以故障指示器的检测为例,结合深度信念网络(DBN)技术实现了自动检测.深度信念网络由于其具有无监督预训练的优势,将其应用于实际系统,对现场的故障指示器视频图像的动作状态进行了分类实验.实验结果表明:深度信念网络分类算法相比于支持向量机(SVM)分类算法和BP分类算法有明显的优势,正确识别率达到了100％.该算法在产品检验的应用中满足生产测试的要求,且降低了人工测试的劳动强度,缓解了视觉疲劳问题.     

Visualization for internet of things: power system and financial network cases

Multimedia Tools and Applications - National critical infrastructure networks, such as banks and industrial control systems (ICSs), can be serious damaged in the event of a security incident....     

Text classification based on deep belief network and softmax regression

In this paper, we propose a novel hybrid text classification model based on deep belief network and softmax regression. To solve the sparse high-dimensional matrix computation problem of texts data, a deep belief network is introduced. After the feature extraction with DBN, softmax regression is employed to classify the text in the learned feature space. In pre-training procedures, the deep belief network and softmax regression are first trained, respectively. Then, in the fine-tuning stage, they are transformed into a coherent whole and the system parameters are optimized with Limited-memory Broyden–Fletcher–Goldfarb–Shanno algorithm. The experimental results on Reuters-21,578 and 20-Newsgroup corpus show that the proposed model can converge at fine-tuning stage and perform significantly better than the classical algorithms, such as SVM and KNN.

     

Smart city information processing under internet of things and cloud computing

The Journal of Supercomputing - This study is to explore the smart city information (SCI) processing technology based on the Internet of Things (IoT) and cloud computing, promoting the construction...     

A-DBNF: adaptive deep belief network framework for regression and classification tasks

Applied Intelligence - Many machine learning methods and models have been proposed for multivariate data regression and classification in recent years. Most of them are supervised learning methods,...     

Development and application of internet of things educational tool based on peer to peer network

With growing interest in IoT across the Peer-to-Peer network industry, basic education programs on IoT for learners are expanding. The purpose of this study was to analyze the difficulties met by learners in basic IoT education, to design and develop hardware tool, and to evaluate the usability of these tool. In this study, the needs of teachers and learners on the tools were analyzed, and design guidelines established. The design guidelines, based on the analysis of needs, suggest the functions of hardware board, miniaturization of board, modularization of sensor parts, and usability and expandability of the tool. A hardware board is designed and developed based on the guidelines. The developed board is then used for basic education on IoT for learners, and the usability of the board is evaluated. The results showed that the new hardware tools provide higher usability than the existing Arduino board. This study is meaningful in that it provides a reference for future IoT educational tools, and that it highlights the importance of promoting learner-centered education.     

An efficient networking protocol for internet of things to handle multimedia big data

In recent years, the emergence of multimedia big data (MBD) due to the excessive use of mobile Internet of Things (ioT) is imposing various challenges to develop efficient communication with the digital world. In this aspect, Mobile Adhoc Network based IoT (MANET-IoT) system is becoming popular due to its greater mobility support and cost-effective nature. A mobile ad hoc network (MANET) consists of randomly placed, battery-powered, moving nodes without an infrastructure that can administer and control traffic in the IoT network. In the MANET-IoT network, the major problems include energy consumption and congestion control to handle MBD data. In this paper, we present two proposals for solving these problems. In the first proposal, a new clustering approach that depends on a well-known protocol called the Low Energy Adaptive Clustering Hierarchy (LEACH) been used in a wireless sensor network (WSN) with modification to adapt to the MANET-IoT’s mobility. Our proposal for applying LEACH to a MANET-IoT consists of rounds, each containing three ordered phases as follows: (1) the announcement phase, in which all nodes announce their remaining energy and the node with the original message also announces itself; (2) the setup phase, in which all cluster heads are selected based on the probability factor with a cycling method; and (3) the steady state phase, in which message delivery to all nodes occurs using several types of links. The second proposal is to provide congestion control for all mobile nodes by link utilization that can support different data rates depending on the link status. Simulation results comparing our modified LEACH protocol to state-of-the-art protocols with utilized links show a great enhancement in energy consumption, received data, throughput, and delay.