Synthetic aperture radar (SAR) image classification is a popular yet challenging research topic in the field of SAR image interpretation. This paper presents a new classification method based on extreme learning machine (ELM) and the superpixel-guided composite kernels (SGCK). By introducing the generalized likelihood ratio (GLR) similarity, a modified simple linear iterative clustering (SLIC) algorithm is firstly developed to generate superpixel for SAR image. Instead of using a fixed-size region, the shape-adaptive superpixel is used to exploit the spatial information, which is effective to classify the pixels in the detailed and near-edge regions. Following the framework of composite kernels, the SGCK is constructed base on the spatial information and backscatter intensity information. Finally, the SGCK is incorporated an ELM classifier. Experimental results on both simulated SAR image and real SAR image demonstrate that the proposed framework is superior to some traditional classification methods.

New deblocking artifact, or blocking artifact reduction, algorithms based on nonlinear adaptive soft-threshold anisotropic filter in wavelet are proposed. Our deblocking algorithm uses soft-threshold, adaptive wavelet direction, adaptive anisotropic filter, and estimation. The novelties of this paper are an adaptive soft-threshold for deblocking artifact and an optimal intersection of confidence intervals (OICI) method in deblocking artifact estimation. The soft-threshold values are adaptable to different thresholds of flat area, texture area, and blocking artifact. The OICI is a reconstruction technique of estimated deblocking artifact which improves acceptable quality level of estimated deblocking artifact and reduces execution time of deblocking artifact estimation compared to the other methods. Our adaptive OICI method outperforms other adaptive deblocking artifact methods. Our estimated deblocking artifact algorithms have up to 98% of MSE improvement, up to 89% of RMSE improvement, and up to 99% of MAE improvement. We also got up to 77.98% reduction of computational time of deblocking artifact estimations, compared to other methods. We have estimated shift and add algorithms by using Euler++(E++) and Runge-Kutta of order 4++ (RK4++) algorithms which iterate one step an ordinary differential equation integration method. Experimental results showed that our E++ and RK4++ algorithms could reduce computational time in terms of shift and add, and RK4++ algorithm is superior to E++ algorithm.

Bandwidth reservation is an important way to guarantee deterministic end-to-end service quality. However, with the traditional bandwidth reservation mechanism, the allocated bandwidth at each link is by default the same without considering the available resource of each link, which may lead to unbalanced resource utilization and limit the number of user connections that network can accommodate. In this paper, we propose a non-uniform bandwidth reservation method, which can further balance the resource utilization of network by optimizing the reserved bandwidth at each link according to its link load. Furthermore, to implement the proposed method, we devise a flexible and automatic bandwidth reservation mechanism based on meter table of Openflow. Through simulations, it is showed that our method can achieve better load balancing performance and make network accommodate more user connections comparing with the traditional methods in most application scenarios.

The traditional tests are planned and designed at the early stages, but it is possible to execute test cases after implementing source code. Since there is a time difference between design stage and testing stage, by the time a software design error is found it will be too late. To solve this problem, this paper suggests a virtual pre-testing process. While the virtual pre-testing process can find software and testing errors before the developing stage, it can automatically generate and execute test cases with modeling and simulation (M&S) in a virtual environment. The first part of this method is to create test cases with state transition tree based on state diagram, which include state, transition, instruction pair, and all path coverage. The second part is to model and simulate a virtual target, which then pre-test the target with test cases. In other words, these generated test cases are automatically transformed into the event list. This simultaneously executes test cases to the simulated target within a virtual environment. As a result, it is possible to find the design and test error at the early stages of the development cycle and in turn can reduce development time and cost as much as possible.

We consider online linear optimization over symmetric positive semi-definite matrices, which has various applications including the online collaborative filtering. The problem is formulated as a repeated game between the algorithm and the adversary, where in each round t the algorithm and the adversary choose matrices Xt and Lt, respectively, and then the algorithm suffers a loss given by the Frobenius inner product of Xt and Lt. The goal of the algorithm is to minimize the cumulative loss. We can employ a standard framework called Follow the Regularized Leader (FTRL) for designing algorithms, where we need to choose an appropriate regularization function to obtain a good performance guarantee. We show that the log-determinant regularization works better than other popular regularization functions in the case where the loss matrices Lt are all sparse. Using this property, we show that our algorithm achieves an optimal performance guarantee for the online collaborative filtering. The technical contribution of the paper is to develop a new technique of deriving performance bounds by exploiting the property of strong convexity of the log-determinant with respect to the loss matrices, while in the previous analysis the strong convexity is defined with respect to a norm. Intuitively, skipping the norm analysis results in the improved bound. Moreover, we apply our method to online linear optimization over vectors and show that the FTRL with the Burg entropy regularizer, which is the analogue of the log-determinant regularizer in the vector case, works well.

We have previously introduced the static dependency pair method that proves termination by analyzing the static recursive structure of various extensions of term rewriting systems for handling higher-order functions. The key is to succeed with the formalization of recursive structures based on the notion of strong computability, which is introduced for the termination of typed λ-calculi. To bring the static dependency pair method close to existing functional programs, we also extend the method to term rewriting models in which functional abstractions with patterns are permitted. Since the static dependency pair method is not sound in general, we formulate a class; namely, accessibility, in which the method works well. The static dependency pair method is a very natural reasoning; therefore, our extension differs only slightly from previous results. On the other hand, a soundness proof is dramatically difficult.

Network function virtualization improves the flexibility of infrastructure resource allocation but the application of commodity facilities arouses new challenges for systematic reliability. To meet the carrier-class reliability demanded from the 5G mobile core, several studies have tackled backup schemes for the virtual network function deployment. However, the existing backup schemes usually sacrifice the efficiency of resource allocation and prevent the sharing of infrastructure resources. To solve the dilemma of balancing the high level demands of reliability and resource allocation in mobile networks, this paper proposes an approach for the problem of pooling deployment of virtualized network functions in virtual EPC network. First, taking pooling of VNFs into account, we design a virtual network topology for virtual EPC. Second, a node-splitting algorithm is proposed to make best use of substrate network resources. Finally, we realize the dynamic adjustment of pooling across different domains. Compared to the conventional virtual topology design and mapping method (JTDM), this approach can achieve fine-grained management and overall scheduling of node resources; guarantee systematic reliability and optimize global view of network. It is proven by a network topology instance provided by SNDlib that the approach can reduce total resource cost of the virtual network and increase the ratio of request acceptance while satisfy the high-demand reliability of the system.

The tens of billions of devices expected to be connected to the Internet will include so many sensors that the demand for sensor-based services is rising. The task of effectively utilizing the enormous numbers of sensors deployed is daunting. The need for automatic sensor identification has expanded the need for research on sensor similarity searches. The Internet of Things (IoT) features massive non-textual dynamic data, which is raising the critical challenge of efficiently and effectively searching for and selecting the sensors most related to a need. Unfortunately, single-attribute similarity searches are highly inaccurate when searching among similar attribute values. In this paper, we propose a group-fitting correlation calculation algorithm (GFC) that can identify the most similar clusters of sensors. The GFC method considers multiple attributes (e.g., humidity, temperature) to calculate sensor similarity; thus, it performs more accurate searches than do existing solutions.

Efficiently locating nodes and allocating demand has been a significant problem for telecommunication network carriers. Most of location models focused on where to locate nodes and how to assign increasing demand with optical access networks. However, the population in industrialized countries will decline over the coming decades. Recent advance in the optical amplifier technology has enabled node integration; an excess telecommunication node is closed and integrated to another node. Node integration in low-demand areas will improve the efficiency of access networks in this approaching age of depopulation. A dynamic node integration problem (DNIP) has been developed to organize the optimal plan for node integration. The problem of the DNIP was that it cannot consider the requirements of network carriers. In actual situations, network carriers often want to specify the way each node is managed, regardless of the mathematical optimality of the solution. This paper proposes a requirement modeling language (RML) for the DNIP, with which the requirements of network carriers can be described. The described statements are used to solve the DNIP, and consequently the calculated optimal solution always satisfies the requirements. The validity of the proposed method was evaluated with computer simulations in a case study.

Automatic speech recognition (ASR) and keyword search (KWS) have more and more found their way into our everyday lives, and their successes could boil down lots of factors. In these factors, large scale of speech data used for acoustic modeling is the key factor. However, it is difficult and time-consuming to acquire large scale of transcribed speech data for some languages, especially for low-resource languages. Thus, at low-resource condition, it becomes important with which transcribed data for acoustic modeling for improving the performance of ASR and KWS. In view of using acoustic data for acoustic modeling, there are two different ways. One is using the target language data, and another is using large scale of other source languages data for cross-lingual transfer. In this paper, we propose some approaches for efficient selecting acoustic data for acoustic modeling. For target language data, a submodular based unsupervised data selection approach is proposed. The submodular based unsupervised data selection could select more informative and representative utterances for manual transcription for acoustic modeling. For other source languages data, the high misclassified as target language based submodular multilingual data selection approach and knowledge based group multilingual data selection approach are proposed. When using selected multilingual data for multilingual deep neural network training for cross-lingual transfer, it could improve the performance of ASR and KWS of target language. When comparing our proposed multilingual data selection approach with language identification based multilingual data selection approach, our proposed approach also obtains better effect. In this paper, we also analyze and compare the language factor and the acoustic factor influence on the performance of ASR and KWS. The influence of different scale of target language data on the performance of ASR and KWS at mono-lingual condition and cross-lingual condition are also compared and analyzed, and some significant conclusions can be concluded.

This paper proposes a novel domain adaptation method that can utilize out-of-domain text resources and partially domain matched text resources in language modeling. A major problem in domain adaptation is that it is hard to obtain adequate adaptation effects from out-of-domain text resources. To tackle the problem, our idea is to carry out model merger in a latent variable space created from latent words language models (LWLMs). The latent variables in the LWLMs are represented as specific words selected from the observed word space, so LWLMs can share a common latent variable space. It enables us to perform flexible mixture modeling with consideration of the latent variable space. This paper presents two types of mixture modeling, i.e., LWLM mixture models and LWLM cross-mixture models. The LWLM mixture models can perform a latent word space mixture modeling to mitigate domain mismatch problem. Furthermore, in the LWLM cross-mixture models, LMs which individually constructed from partially matched text resources are split into two element models, each of which can be subjected to mixture modeling. For the approaches, this paper also describes methods to optimize mixture weights using a validation data set. Experiments show that the mixture in latent word space can achieve performance improvements for both target domain and out-of-domain compared with that in observed word space.

Automatically recognizing pain and estimating pain intensity is an emerging research area that has promising applications in the medical and healthcare field, and this task possesses a crucial role in the diagnosis and treatment of patients who have limited ability to communicate verbally and remains a challenge in pattern recognition. Recently, deep learning has achieved impressive results in many domains. However, deep architectures require a significant amount of labeled data for training, and they may fail to outperform conventional handcrafted features due to insufficient data, which is also the problem faced by pain detection. Furthermore, the latest studies show that handcrafted features may provide complementary information to deep-learned features; hence, combining these features may result in improved performance. Motived by the above considerations, in this paper, we propose an innovative method based on the combination of deep spatiotemporal and handcrafted features for pain intensity estimation. We use C3D, a deep 3-dimensional convolutional network that takes a continuous sequence of video frames as input, to extract spatiotemporal facial features. C3D models the appearance and motion of videos simultaneously. For handcrafted features, we propose extracting the geometric information by computing the distance between normalized facial landmarks per frame and the ones of the mean face shape, and we extract the appearance information using the histogram of oriented gradients (HOG) features around normalized facial landmarks per frame. Two levels of SVRs are trained using spatiotemporal, geometric and appearance features to obtain estimation results. We tested our proposed method on the UNBC-McMaster shoulder pain expression archive database and obtained experimental results that outperform the current state-of-the-art.

In this paper, the problem of lifetime extension of wireless sensor networks (WSNs) with redundant sensor nodes deployed in 3D vegetation-covered fields is modeled, which includes building communication models, network model and energy model. Generally, such a problem cannot be solved by a conventional method directly. Here we propose an Artificial Bee Colony (ABC) based optimal grouping algorithm (ABC-OG) to solve it. The main contribution of the algorithm is to find the optimal number of feasible subsets (FSs) of WSN and assign them to work in rotation. It is verified that reasonably grouping sensors into FSs can average the network energy consumption and prolong the lifetime of the network. In order to further verify the effectiveness of ABC-OG, two other algorithms are included for comparison. The experimental results show that the proposed ABC-OG algorithm provides better optimization performance.

Geosocial networking allows users to interact with respect to their current locations, which enables a group of users to determine where to meet. This calls for techniques that support processing of Multiple-user Location-based Keyword (MULK) queries, which return a set of Point-of-Interests (POIs) that are 'close' to the locations of the users in a group and can provide them with potential options at the lowest expense (e.g., minimizing travel distance). In this paper, we formalize the MULK query and propose a dynamic programming-based algorithm to find the optimal result set. Further, we design two approximation algorithms to improve MULK query processing efficiency. The experimental evaluations show that our solutions are feasible and efficient under various parameter settings.

There is a strong demand for super-resolution time of arrival (TOA) estimation techniques for radar applications that can that can exceed the theoretical limits on range resolution set by frequency bandwidth. One of the most promising solutions is the use of compressed sensing (CS) algorithms, which assume only the sparseness of the target distribution but can achieve super-resolution. To preserve the reconstruction accuracy of CS under highly correlated and noisy conditions, we introduce a random resampling approach to process the received signal and thus reduce the coherent index, where the frequency-domain-based CS algorithm is used as noise reduction preprocessing. Numerical simulations demonstrate that our proposed method can achieve super-resolution TOA estimation performance not possible with conventional CS methods.

Response property is a kind of liveness property. Response property problem is defined as follows: Given two activities α and β, whenever α is executed, is β always executed after that? In this paper, we tackled the problem in terms of Workflow Petri nets (WF-nets for short). Our results are (i) the response property problem for acyclic WF-nets is decidable, (ii) the problem is intractable for acyclic asymmetric choice (AC) WF-nets, and (iii) the problem for acyclic bridge-less well-structured WF-nets is solvable in polynomial time. We illustrated the usefulness of the procedure with an application example.

We present a new action classification method for skeletal sequence data. The proposed method is based on simple nonparametric feature matching without a learning process. We first augment the training dataset to implicitly construct an exponentially increasing number of training sequences, which can be used to improve the generalization power of the proposed action classifier. These augmented training sequences are matched to the test sequence with the relaxed dynamic time warping (DTW) technique. Our relaxed formulation allows the proposed method to work faster and with higher efficiency than the conventional DTW-based method using a non-augmented dataset. Experimental results show that the proposed approach produces effective action classification results for various scales of real datasets.

Radar is expected in advanced driver-assistance systems for environmentally robust measurements. In this paper, we propose a novel radar signal segmentation method by using a complex-valued fully convolutional network (CvFCN) that comprises complex-valued layers, real-valued layers, and a bidirectional conversion layer between them. We also propose an efficient automatic annotation system for dataset generation. We apply the CvFCN to two-dimensional (2D) complex-valued radar signal maps (r-maps) that comprise angle and distance axes. An r-maps is a 2D complex-valued matrix that is generated from raw radar signals by 2D Fourier transformation. We annotate the r-maps automatically using LiDAR measurements. In our experiment, we semantically segment r-map signals into pedestrian and background regions, achieving accuracy of 99.7% for the background and 96.2% for pedestrians.

Damage caused by malware has become a serious problem. The recent rise in the spread of evasive malware has made it difficult to detect it at the pre-infection timing. Malware detection at post-infection timing is a promising approach that fulfills this gap. Given this background, this work aims to identify likely malware-infected devices from the measurement of Internet traffic. The advantage of the traffic-measurement-based approach is that it enables us to monitor a large number of endhosts. If we find an endhost as a source of malicious traffic, the endhost is likely a malware-infected device. Since the majority of malware today makes use of the web as a means to communicate with the C&C servers that reside on the external network, we leverage information recorded in the HTTP headers to discriminate between malicious and benign traffic. To make our approach scalable and robust, we develop the automatic template generation scheme that drastically reduces the amount of information to be kept while achieving the high accuracy of classification; since it does not make use of any domain knowledge, the approach should be robust against changes of malware. We apply several classifiers, which include machine learning algorithms, to the extracted templates and classify traffic into two categories: malicious and benign. Our extensive experiments demonstrate that our approach discriminates between malicious and benign traffic with up to 97.1% precision while maintaining the false positive rate below 1.0%.

Recently, mobile applications for recording everyday meals draw much attention for self dietary. However, most of the applications return food calorie values simply associated with the estimated food categories, or need for users to indicate the rough amount of foods manually. In fact, it has not been achieved to estimate food calorie from a food photo with practical accuracy, and it remains an unsolved problem. Then, in this paper, we propose estimating food calorie from a food photo by simultaneous learning of food calories, categories, ingredients and cooking directions using deep learning. Since there exists a strong correlation between food calories and food categories, ingredients and cooking directions information in general, we expect that simultaneous training of them brings performance boosting compared to independent single training. To this end, we use a multi-task CNN. In addition, in this research, we construct two kinds of datasets that is a dataset of calorie-annotated recipe collected from Japanese recipe sites on the Web and a dataset collected from an American recipe site. In the experiments, we trained both multi-task and single-task CNNs, and compared them. As a result, a multi-task CNN achieved the better performance on both food category estimation and food calorie estimation than single-task CNNs. For the Japanese recipe dataset, by introducing a multi-task CNN, 0.039 were improved on the correlation coefficient, while for the American recipe dataset, 0.090 were raised compared to the result by the single-task CNN. In addition, we showed that the proposed multi-task CNN based method outperformed search-based methods proposed before.