Foreword to the special issue of the intelligent systems for the Internet of Things ( ISIT2018 )

The purpose of this special issue is to assemble a selection of best research articles that were presented at the third annual workshop on Intelligent Systems for the Internet of Things held in Wuhan. A number of practitioners, scholars, researchers, and vendors attended this workshop. In this workshop, some academic ideas, engineering technologies, future collaborations, and new research directions on IoT were discussed. With the advent of the fourth Industrial Revolution, many technologies, including big data analytics, blockchain, artificial intelligence, and data mining, have also been used for exploring and analyzing data generated from the IoT field. In addition to these, intelligent technologies such as colony optimization, particle swarm optimization, and simulated annealing also provide many solutions for the IoT applications. These provided solutions not only can enhance the performance of an IoT system and its devices, but also can make a system aware of events occurred. This special issue presents many excellent works on how scholars, researchers, vendors, and engineers are collaborating to address high-performance edge complicated research challenges. The scope of this special issue is broad and is representative of the multidisciplinary nature of the IoT. In addition to submissions that deal with intelligent technologies for the IoT and their applications, this issue also includes articles that address practical challenges with IoT-related technologies, such as edge computing, cloud computing, blockchain, computer vision, and deep learning technologies. In the article Intelligent cloud computing platform for 3D sound reproduction,1 3D sound reproduction is a hot topic in virtual reality. Both Dolby and DTS pay great attention to 3D sound reproduction research. Although there exist several 3D sound reproduction methods, few techniques are practical. The requirements in state-of-the-art techniques are critical, such as that all the loudspeakers are on a sphere, the calculation is complicated. The authors develop a novel 3D sound reproduction method and all the parameters in the proposed system are generated on a cloud platform. It is convenient to configure loudspeaker array for users and it is also helpful for the popularization of 3D sound reproduction. In the article Speech emotion recognition using emotion perception spectral feature,2 a new speech spectral feature is provided for speech emotion recognition. Speech emotion recognition is an important technique for human–computer interface applications. Due to its rich information of emotion, the spectral feature is widely used for emotion recognition. However, the recognition performance is limited because of the imprecise extracted rule and uncertain size of resolution of the spectral feature. To address this issue, the authors were motivated by speech coding, they introduced psychoacoustics model, and provided a perception spectral subband partition method for obtaining more precise frequency resolution. Moreover, they also provided a new spectral feature on the divided subband frequency signals. The proposed feature includes emotional perception entropy, spectral inclination, and spectral flatness. Then, a support vector machine classifier is used to recognize emotion categories. The experimental results show that the proposed spectral feature is superior to the traditional MFCC feature, and also better than the state-of-the-art Fourier feature and the multiresolution amplitude feature. This article proposed a new ideal for emotion feature extracting and broaden solution of speech emotion recognition. In the article Real-time action feature extraction via fast PCA-Flow,3 a novel real-time action feature extraction method for human action recognition is provided. Video-based human action recognition can be widely used in network video retrieval, video surveillance analysis, medical video monitoring, and so forth. So it has attracted the attention of scholars who majored in image analysis. The difficulty of action recognition researchers is how to improve the accuracy of action features while reducing the complexity of feature calculation. The authors developed a novel real-time video feature extraction technique by exploiting the fast PCA-Flow algorithm. Experimental results indicate that the proposed method properly balances the accuracy of action features and the time of feature computation. In the article A novel infant cry recognition system using auditory model-based robust feature and GMM-UBM,4 the authors provided a novel infant cry recognition system. Recognizing infant cry is a meaningful work, which can help a new parent understand an infant’s needs. Mostly, the motivation of existed recognition features is based on the psychoacoustic model. However, weak features are not enough to represent the details of infant cry. To address this issue, the authors propose a novel infant cry recognition system. In their system, the feature extraction method was derived from the auditory model. This model can address the auditory neural active representation. In addition, they designed the intelligent system by a classifier named Gaussian mixture model-universal background model, which has a robust recognition performance to channel imbalance and corroded signal. The experiment results have shown a superior performance. Compared with the high-order spectral feature, a state-of-the-art feature, when using Gaussian mixture model and general background model as the classifier, the recognition accuracy improved by 14.6% and 13.7% under clear and corroded signals, respectively. This article on infant cry recognition is an interesting work.