Neal N. Xiong

Bio: Neal N. Xiong is an academic researcher from Jiangxi University of Finance and Economics. The author has contributed to research in topics: Gross domestic product & Identification (biology). The author has an hindex of 3, co-authored 3 publications receiving 57 citations.

Optimized Multioperator Image Retargeting Based on Perceptual Similarity Measure

Abstract: With various emerging mobile devices, the visual content have be to resized into different sizes or aspect ratios for good viewing experiences. In this paper, we propose a new multioperator retargeting algorithm by using four retargeting operators of seam carving, cropping, warping, and scaling iteratively. To determine which retargeting operator should be used at each iteration, we adopt structural similarity (SSIM) to evaluate the similarity between the original and retargeted images. The retargeting operator sequence is constructed based on the four types of retargeting operators by an optimization process. Since the sizes of original and retargeted images are different, scale-invariant feature transform flow is used for dense correspondence between the original and retargeted images for similarity evaluation. Additionally, visual saliency is used to weight SSIM results based on the characteristics of the human visual system. Experimental results on a public image retargeting database have shown the promising performance of the proposed multioperator retargeting algorithm.

A Secure, Intelligent, and Smart-Sensing Approach for Industrial System Automation and Transmission over Unsecured Wireless Networks

Abstract: In Industrial systems, Supervisory control and data acquisition (SCADA) system, the pseudo-transport layer of the distributed network protocol (DNP3) performs the functions of the transport layer and network layer of the open systems interconnection (OSI) model. This study used a simulation design of water pumping system, in-which the network nodes are directly and wirelessly connected with sensors, and are monitored by the main controller, as part of the wireless SCADA system. This study also intends to focus on the security issues inherent in the pseudo-transport layer of the DNP3 protocol. During disassembly and reassembling processes, the pseudo-transport layer keeps track of the bytes sequence. However, no mechanism is available that can verify the message or maintain the integrity of the bytes in the bytes received/transmitted from/to the data link layer or in the send/respond from the main controller/sensors. To properly and sequentially keep track of the bytes, a mechanism is required that can perform verification while bytes are received/transmitted from/to the lower layer of the DNP3 protocol or the send/respond to/from field sensors. For security and byte verification purposes, a mechanism needs to be proposed for the pseudo-transport layer, by employing cryptography algorithm. A dynamic choice security buffer (SB) is designed and employed during the security development. To achieve the desired goals of the proposed study, a pseudo-transport layer stack model is designed using the DNP3 protocol open library and the security is deployed and tested, without changing the original design.

The Study on Stage Financing Model of IT Project Investment

Abstract: Stage financing is the basic operation of venture capital investment. In investment, usually venture capitalists use different strategies to obtain the maximum returns. Due to its advantages to reduce the information asymmetry and agency cost, stage financing is widely used by venture capitalists. Although considerable attentions are devoted to stage financing, very little is known about the risk aversion strategies of IT projects. This paper mainly addresses the problem of risk aversion of venture capital investment in IT projects. Based on the analysis of characteristics of venture capital investment of IT projects, this paper introduces a real option pricing model to measure the value brought by the stage financing strategy and design a risk aversion model for IT projects. Because real option pricing method regards investment activity as contingent decision, it helps to make judgment on the management flexibility of IT projects and then make a more reasonable evaluation about the IT programs. Lastly by being applied to a real case, it further illustrates the effectiveness and feasibility of the model.

Multi-Center Agent Loss for Visual Identification of Chinese Simmental in the Wild

Abstract: Simple Summary Visual identification of cattle in a realistic farming environment is helpful for real-time cattle monitoring. Based on continuous cattle detection, identification, and behavior recognition, it is possible to utilize cameras on farms within company or government networks to provide the services of production supervision, early disease detection, and animal science research for precision livestock farming. However, cattle identification in the wild is still a difficult problem due to the high similarities of different identities and the variances of the same identity as posture or perspective changes. Our proposed method based on deep convolutional neural networks and deep metric learning provides a promising approach for cattle identification and paves the way toward continuous monitoring of cattle in a nearly natural state. Abstract Visual identification of cattle in the wild provides an essential way for real-time cattle monitoring applicable to precision livestock farming. Chinese Simmental exhibit a yellow or brown coat with individually characteristic white stripes or spots, which makes a biometric identifier for identification possible. This work employed the observable biometric characteristics to perform cattle identification with an image from any viewpoint. We propose multi-center agent loss to jointly supervise the learning of DCNNs by SoftMax with multiple centers and the agent triplet. We reformulated SoftMax with multiple centers to reduce intra-class variance by offering more centers for feature clustering. Then, we utilized the agent triplet, which consisted of the features and the agents, to enforce separation among different classes. As there are no datasets for the identification of cattle with multi-view images, we created CNSID100, consisting of 11,635 images from 100 Chinese Simmental identities. Our proposed loss was comprehensively compared with several well-known losses on CNSID100 and OpenCows2020 and analyzed in an engineering application in the farming environment. It was encouraging to find that our approach outperformed the state-of-the-art models on the datasets above. The engineering application demonstrated that our pipeline with detection and recognition is promising for continuous cattle identification in real livestock farming scenarios.

Human Action Monitoring for Healthcare Based on Deep Learning

Abstract: Human action monitoring can be advantageous to remotely monitor the status of patients or elderly person for intelligent healthcare Human action recognition enables efficient and accurate monitoring of human behaviors, which can exhibit multifaceted complexity attributed to disparities in viewpoints, personality, resolution and motion speed of individuals, etc The spatial-temporal information plays an important role in the human action recognition In this paper, we proposed a novel deep learning architecture named as recurrent 3D convolutional neural network (R3D) to extract effective and discriminative spatial-temporal features to be used for action recognition, which enables the capturing of long-range temporal information by aggregating the 3D convolutional network entries to serve as an input to the LSTM (Long Short-Term Memory) architecture The 3D convolutional network and LSTM are two effective methods for extracting the temporal information The proposed R3D network integrated these two methods by sharing a shared 3D convolutional network in sliding windows on video streaming to capturing short-term spatial-temporal features into the LSTM The output features of LSTM encapsulate the long-range spatial-temporal information representing high-level abstraction of the human actions The proposed algorithm is compared to traditional and the-state-of-the-art and deep learning algorithms The experimental results demonstrated the effectiveness of the proposed system, which can be used as smart monitoring for remote healthcare

The Influencing Factors of Enterprise Sustainable Innovation: An Empirical Study

Abstract: Sustainable innovation is the inexhaustible source of development of enterprises. Within fierce market competition, only by depending on continuous innovation can an enterprise exist and develop. By conducting an exploratory factor analysis and a confirmatory factor analysis, this paper proposes a theoretical model, dividing enterprise sustainable innovation ability into three aspects: knowledge innovation capability, production innovation capability, and market innovation capability, and analyzes the influencing factors respectively. Finally, applying this theoretical model to a practical case, with system dynamics method, the simulation results show that they are consistent with real enterprise facts. Therefore, the framework of determinants of sustainable innovation built in this paper has already been verified theoretically and practically. It not only lays a theoretical foundation for further research, but also provides a clear ground for firms to improve their sustainable innovation.

Copy-move forgery detection using combined features and transitive matching

Abstract: Recently, the research of Internet of Things (IoT) and Multimedia Big Data (MBD) has been growing tremendously. Both IoT and MBD have a lot of multimedia data, which can be tampered easily. Therefore, the research of multimedia forensics is necessary. Copy-move is an important branch of multimedia forensics. In this paper, a novel copy-move forgery detection scheme using combined features and transitive matching is proposed. First, SIFT and LIOP are extracted as combined features from the input image. Second, transitive matching is used to improve the matching relationship. Third, a filtering approach using image segmentation is proposed to filter out false matches. Fourth, affine transformations are estimated between these image patches. Finally, duplicated regions are located based on those affine transformations. The experimental results demonstrate that the proposed scheme can achieve much better detection results on the public database under various attacks.

Learning a Distance Metric by Balancing KL-Divergence for Imbalanced Datasets

Abstract: In many real-world domains, datasets with imbalanced class distributions occur frequently, which may confuse various machine learning tasks. Among all these tasks, learning classifiers from imbalanced datasets is an important topic. To perform this task well, it is crucial to train a distance metric which can accurately measure similarities between samples from imbalanced datasets. Unfortunately, existing distance metric methods, such as large margin nearest neighbor, information-theoretic metric learning, etc., care more about distances between samples and fail to take imbalanced class distributions into consideration. Traditional distance metrics have natural tendencies to favor the majority classes, which can more easily satisfy their objective function. Those important minority classes are always neglected during the construction process of distance metrics, which severely affects the decision system of most classifiers. Therefore, how to learn an appropriate distance metric which can deal with imbalanced datasets is of vital importance, but challenging. In order to solve this problem, this paper proposes a novel distance metric learning method named distance metric by balancing KL-divergence (DMBK). DMBK defines normalized divergences using KL-divergence to describe distinctions between different classes. Then it combines geometric mean with normalized divergences and separates samples from different classes simultaneously. This procedure separates all classes in a balanced way and avoids inaccurate similarities incurred by imbalanced class distributions. Various experiments on imbalanced datasets have verified the excellent performance of our novel method.