Due to its simplicity and efficiency, differential evolution (DE) has gained the interest of researchers from various fields for solving global optimization problems. However, it is prone to premature convergence at local minima. To overcome this drawback, a novel hybrid dragonfly algorithm with differential evolution (Hybrid DA-DE) for solving global optimization problems is proposed. Firstly, a novel mutation operator is introduced based on the dragonfly algorithm (DA). Secondly, the scaling factor (F) is adjusted in a self-adaptive and individual-dependent way without extra parameters. The proposed algorithm combines the exploitation capability of DE and exploration capability of DA to achieve optimal global solutions. The effectiveness of this algorithm is evaluated using 30 classical benchmark functions with sixteen state-of-the-art meta-heuristic algorithms. A series of experimental results show that Hybrid DA-DE outperforms other algorithms significantly. Meanwhile, Hybrid DA-DE has the best adaptability to high-dimensional problems.

A triangle enumerating problem is one of fundamental problems of graph data. Although several triangle enumerating algorithms based on MapReduce have been proposed, they still suffer from generating a lot of intermediate data. In this paper, we propose the efficient MapReduce algorithms to enumerate every triangle in the massive graph based on a vertex partition. Since a triangle is composed of an edge and a wedge, our algorithms check the existence of an edge connecting the end-nodes of each wedge. To generate every triangle from a graph in parallel, we first split a graph into several vertex partitions and group the edges and wedges in the graph for each pair of vertex partitions. Then, we form the triangles appearing in each group. Furthermore, to enhance the performance of our algorithm, we remove the duplicated wedges existing in several groups. Our experimental evaluation shows the performance of our proposed algorithm is better than that of the state-of-the-art algorithm in diverse environments.

The Cuckoo Search (CS) is apt to be trapped in local optimum relating to complex target functions. This drawback has been recognized as the bottleneck of its widespread use. This paper, with the purpose of improving CS, puts forward a Cuckoo Search algorithm featuring Multi-Learning Strategies (LSCS). In LSCS, the Converted Learning Module, which features the Comprehensive Learning Strategy and Optimal Learning Strategy, tries to make a coordinated cooperation between exploration and exploitation, and the switching in this part is decided by the transition probability P_c. When the nest fails to be renewed after m iterations, the Elite Learning Perturbation Module provides extra diversity for the current nest, and it can avoid stagnation. The Boundary Handling Approach adjusted by Gauss map is utilized to reset the location of nest beyond the boundary. The proposed algorithm is evaluated by two different tests: Test Group A(ten simple unimodal and multimodal functions) and Test Group B(the CEC2013 test suite). Experiments results show that LSCS demonstrates significant advantages in terms of convergence speed and optimization capability in solving complex problems.

We design a new oblivious routing algorithm for two-dimensional mesh-based Networks-on-Chip (NoCs) called LEF (Long Edge First) which offers high throughput with low design complexity. LEF's basic idea comes from conventional wisdom in choosing the appropriate dimension-order routing (DOR) algorithm for supercomputers with asymmetric mesh or torus interconnects: routing longest dimensions first provides better performance than other strategies. In LEF, we combine the XY DOR and the YX DOR. When routing a packet, which DOR algorithm is chosen depends on the relative position between the source node and the destination node. Decisions of selecting the appropriate DOR algorithm are not fixed to the network shape but instead made on a per-packet basis. We also propose an efficient deadlock avoidance method for LEF in which the use of virtual channels is more flexible than in the conventional method. We evaluate LEF against O1TURN, another effective oblivious routing algorithm, and a minimal adaptive routing algorithm based on the odd-even turn model. The evaluation results show that LEF is particularly effective when the communication is within an asymmetric mesh. In a 16×8 NoC, LEF even outperforms the adaptive routing algorithm in some cases and delivers from around 4% up to around 64.5% higher throughput than O1TURN. Our results also show that the proposed deadlock avoidance method helps to improve LEF's performance significantly and can be used to improve O1TURN's performance. We also examine LEF in large-scale NoCs with thousands of nodes. Our results show that, as the NoC size increases, the performance of the routing algorithms becomes more strongly influenced by the resource allocation policy in the network and the effect is different for each algorithm. This is evident in that results of middle-scale NoCs with around 100 nodes cannot be applied directly to large-scale NoCs.

For cloud data center, Virtual Machine (VM) consolidation is an effective way to save energy and improve efficiency. However, inappropriate consolidation of VMs, especially aggressive consolidation, can lead to performance problems, and even more serious Service Level Agreement (SLA) violations. Therefore, it is very important to solve the tradeoff between reduction in energy use and reduction of SLA violation level. In this paper, we propose two Host State Detection algorithms and an improved VM placement algorithm based on our proposed Host State Binary Decision Tree Prediction model for SLA-aware and energy-efficient consolidation of VMs in cloud data centers. We propose two formulas of conditions for host state estimate, and our model uses them to build a Binary Decision Tree manually for host state detection. We extend Cloudsim simulator to evaluate our algorithms by using PlanetLab workload and random workload. The experimental results show that our proposed model can significantly reduce SLA violation rates while keeping energy cost efficient, it can reduce the metric of SLAV by at most 98.12% and the metric of Energy by at most 33.96% for real world workload.

A program is non-interferent if it leaks no secret information to an observable output. However, non-interference is too strict in many practical cases and quantitative information flow (QIF) has been proposed and studied in depth. Originally, QIF is defined as the average of leakage amount of secret information over all executions of a program. However, a vulnerable program that has executions leaking the whole secret but has the small average leakage could be considered as secure. This counter-intuition raises a need for a new definition of information leakage of a particular run, i.e., dynamic leakage. As discussed in [5], entropy-based definitions do not work well for quantifying information leakage dynamically; Belief-based definition on the other hand is appropriate for deterministic programs, however, it is not appropriate for probabilistic ones.

In this paper, we propose new simple notions of dynamic leakage based on entropy which are compatible with existing QIF definitions for deterministic programs, and yet reasonable for probabilistic programs in the sense of [5]. We also investigated the complexity of computing the proposed dynamic leakage for three classes of Boolean programs. We also implemented a tool for QIF calculation using model counting tools for Boolean formulae. Experimental results on popular benchmarks of QIF research show the flexibility of our framework. Finally, we discuss the improvement of performance and scalability of the proposed method as well as an extension to more general cases.

Fault prediction aims to identify whether a software module is defect-prone or not according to metrics that are mined from software projects. These metric values, also known as features, may involve irrelevance and redundancy, which hurt the performance of fault prediction models. In order to filter out irrelevant and redundant features, a Hybrid Feature Selection (abbreviated as HFS) method for software fault prediction is proposed. The proposed HFS method consists of two major stages. First, HFS groups features with hierarchical agglomerative clustering; second, HFS selects the most valuable features from each cluster to remove irrelevant and redundant ones based on two wrapper based strategies. The empirical evaluation was conducted on 11 widely-studied NASA projects, using three different classifiers with four performance metrics (precision, recall, F-measure, and AUC). Comparison with six filter-based feature selection methods demonstrates that HFS achieves higher average F-measure and AUC values. Compared with two classic wrapper feature selection methods, HFS can obtain a competitive prediction performance in terms of average AUC while significantly reducing the computation cost of the wrapper process.

Zero-effort bilateral authentication was introduced recently to use a trusted wristwear to continuously authenticate a smartphone user. A user is allowed to use the smartphone if both wristwear and smartphone are determined to be held by the same person by comparing the wristwear's motion with the smartphone's input or motion, depending on the grip — which hand holds the smartphone and which hand provides the input. Unfortunately, the scheme has several shortcomings. First, it may work improperly when the user is walking since the gait can conceal the wrist's motions of making touches. Second, it continuously compares the motions of the two devices, which incurs a heavy communication burden. Third, the acceleration-based grip inference, which assumes that the smartphone is horizontal with the ground is inapplicable in practice. To address these shortcomings, we propose WearAuth, wristwear-assisted user authentication for smartphones in this paper. WearAuth applies wavelet-based multi-resolution analysis to extract the desired touch-specific movements regardless of whether the user is stationary or moving; uses discrete Fourier transform-based approximate correlation to reduce the communication overhead; and takes a new approach to directly compute the relative device orientation without using acceleration to infer the grip more precisely. In two experiments with 50 subjects, WearAuth produced false negative rates of 3.6% or less and false positive rates of 1.69% or less. We conclude that WearAuth operates properly under various usage cases and is robust to sophisticated attacks.

Software diversity can be utilized in cyberspace security to defend against the zero-day attacks. Existing researches have proved the effectiveness of diversity in bringing security benefits, but few of them touch the problem that whether there is a positive correlation between the security and the diversity. In addition, there is little guidance on how to construct an effective diversified system. For that, this paper develops two diversity metrics based on system attribute matrix, proposes a diversity measurement and verifies the effectiveness of the measurement. Through several simulations on the diversified systems which use voting strategy, the relationship between diversity and security is analyzed. The results show that there is an overall positive correlation between security and diversity. Though some cases are against the correlation, further analysis is made to explain the phenomenon. In addition, the effect of voting strategy is also discussed through simulations. The results show that the voting strategy have a dominant impact on the security, which implies that security benefits can be obtained only with proper strategies. According to the conclusions, some guidance is provided in constructing a more diversified as well as securer system.

Recognizing the different segments of speech belonging to the same speaker is an important speech analysis task in various applications. Recent works have shown that there was an underlying manifold on which speaker utterances live in the model-parameter space. However, most speaker clustering methods work on the Euclidean space, and hence often fail to discover the intrinsic geometrical structure of the data space and fail to use such kind of features. For this problem, we consider to convert the speaker i-vector representation of utterances in the Euclidean space into a network structure constructed based on the local (k) nearest neighbor relationship of these signals. We then propose an efficient community detection model on the speaker content network for clustering signals. The new model is based on the probabilistic community memberships, and is further refined with the idea that: if two connected nodes have a high similarity, their community membership distributions in the model should be made close. This refinement enhances the local invariance assumption, and thus better respects the structure of the underlying manifold than the existing community detection methods. Some experiments are conducted on graphs built from two Chinese speech databases and a NIST 2008 Speaker Recognition Evaluations (SREs). The results provided the insight into the structure of the speakers present in the data and also confirmed the effectiveness of the proposed new method. Our new method yields better performance compared to with the other state-of-the-art clustering algorithms. Metrics for constructing speaker content graph is also discussed.

Many quality metrics have been proposed for the compliance perception to assess haptic device performance and perceived results. Perceived compliance may be influenced by factors such as object properties, experimental conditions and human perceptual habits. In this paper, analysis of softness perception was conducted to find out relevant quality metrics dominating in the compliance perception system and their correlation with perception results, by expressing these metrics by basic physical parameters that characterizing these factors. Based on three psychophysical experiments, just noticeable differences (JNDs) for perceived softness of combination of different stiffness coefficients and damping levels rendered by haptic devices were analyzed. Interaction data during the interaction process were recorded and analyzed. Preliminary experimental results show that the discrimination ability of softness perception changes with the ratio of damping to stiffness when subjects exploring at their habitual speed. Analysis results indicate that quality metrics of Rate-hardness, Extended Rate-hardness and ratio of damping to stiffness have high correlation for perceived results. Further analysis results show that parameters that reflecting object properties (stiffness, damping), experimental conditions (force bandwidth) and human perceptual habits (initial speed, maximum force change rate) lead to the change of these quality metrics, which then bring different perceptual feeling and finally result in the change of discrimination ability. Findings in this paper may provide a better understanding of softness perception and useful guidance in improvement of haptic and teleoperation devices.

Edge-preserving smoothing filter smoothes the textures while preserving the information of sharp edges. In image processing, this kind of filter is used as a fundamental process of many applications. In this paper, we propose a new approach for edge-preserving smoothing filter. Our method uses 2D local filter to smooth images and we apply indicator function to restrict the range of filtered pixels for edge-preserving. To define the indicator function, we recalculate the distance between each pixel by using edge information. The nearby pixels in the new domain are used for smoothing. Since our method constrains the pixels used for filtering, its running time is quite fast. We demonstrate the usefulness of our new edge-preserving smoothing method for some applications.

In recent decade, many state-of-the-art algorithms on image classification as well as audio classification have achieved noticeable successes with the development of deep convolutional neural network (CNN). However, most of the works only exploit single type of training data. In this paper, we present a study on classifying bird species by exploiting the combination of both visual (images) and audio (sounds) data using CNN, which has been sparsely treated so far. Specifically, we propose CNN-based multimodal learning models in three types of fusion strategies (early, middle, late) to settle the issues of combining training data cross domains. The advantage of our proposed method lies on the fact that we can utilize CNN not only to extract features from image and audio data (spectrogram) but also to combine the features across modalities. In the experiment, we train and evaluate the network structure on a comprehensive CUB-200-2011 standard data set combing our originally collected audio data set with respect to the data species. We observe that a model which utilizes the combination of both data outperforms models trained with only an either type of data. We also show that transfer learning can significantly increase the classification performance.

Conventional TLB (Translation Lookaside Buffer) coalescing schemes do not fully exploit the contiguity that a memory allocator provides. The conventional schemes accordingly have certain performance overheads due to page table walks. To address this issue, we propose an efficient scheme, called block contiguity translation (BCT), that accommodates the block size information in a page table considering the Buddy algorithm. By fully exploiting the block-level contiguity, we can reduce the page table walks as certain physical memory is allocated in the contiguous way. Additionally, we present unified per-level page sizes to simplify the design and better utilize the contiguity information. Considering the state-of-the-art schemes as references, the comparative analysis and the performance simulations are conducted. Experiments indicate that the proposed scheme can improve the memory system performance with moderate hardware overheads.

This paper proposes a deep neural network (DNN) based framework to address the problem of vector quantization (VQ) for high-dimensional data. The main challenge of applying DNN to VQ is how to reduce the binary coding error of the auto-encoder when the distribution of the coding units is far from binary. To address this problem, three fine-tuning methods have been adopted: 1) adding Gaussian noise to the input of the coding layer, 2) forcing the output of the coding layer to be binary, 3) adding a non-binary penalty term to the loss function. These fine-tuning methods have been extensively evaluated on quantizing speech magnitude spectra. The results demonstrated that each of the methods is useful for improving the coding performance. When implemented for quantizing 968-dimensional speech spectra using only 18-bit, the DNN-based VQ framework achieved an averaged PESQ of about 2.09, which is far beyond the capability of conventional VQ methods.

This study designs a low-cost portable device that functions as a coaching assistant system which can support table tennis practice. Although deep learning technology is a promising solution to realizing human activity recognition, we propose using cosine similarity in making inferences. Our experiments show that the cosine similarity based inference can be a good alternative to the deep learning based inference for the assistant system when resources are limited.

Sound event detection is intended to identify the sound events in audio recordings, which has widespread applications in real life. Recently, convolutional recurrent neural network (CRNN) models have achieved state-of-the-art performance in this task due to their capabilities in learning the representative features. However, the CRNN models are of high complexities with millions of parameters to be trained, which limits their usage for the mobile and embedded devices with limited computation resource. Model distillation is effective to distill the knowledge of a complex model to a smaller one, which can be deployed on the devices with limited computational power. In this letter, we propose a novel multi model-based distillation approach for sound event detection by making use of the knowledge from models of multiple teachers which are complementary in detecting sound events. Extensive experimental results demonstrated that our approach achieves a compression ratio about 50 times. In addition, better performance is obtained for the sound event detection task.

To the best of our knowledge, there are a few researches on air-handwriting character-level writer identification only employing acceleration and angular velocity data. In this paper, we propose a deep learning approach to writer identification only using inertial sensor data of air-handwriting. In particular, we separate different representations of degree of freedom (DoF) of air-handwriting to extract local dependency and interrelationship in different CNNs separately. Experiments on a public dataset achieve an average good performance without any extra hand-designed feature extractions.

The effectiveness of model adaptation in dialogue speech synthesis is explored. The proposed adaptation method is based on a conversion from a base model learned with a large dataset into a target, dialogue-style speech model. The proposed method is shown to improve the intelligibility of synthesized dialogue speech, while maintaining the speaking style of dialogue.

Human gender recognition in the wild is a challenging task due to complex face variations, such as poses, lighting, occlusions, etc. In this letter, learnable Gabor convolutional network (LGCN), a new neural network computing framework for gender recognition was proposed. In LGCN, a learnable Gabor filter (LGF) is introduced and combined with the convolutional neural network (CNN). Specifically, the proposed framework is constructed by replacing some first layer convolutional kernels of a standard CNN with LGFs. Here, LGFs learn intrinsic parameters by using standard back propagation method, so that the values of those parameters are no longer fixed by experience as traditional methods, but can be modified by self-learning automatically. In addition, the performance of LGCN in gender recognition is further improved by applying a proposed feature combination strategy. The experimental results demonstrate that, compared to the standard CNNs with identical network architecture, our approach achieves better performance on three challenging public datasets without introducing any sacrifice in parameter size.

As a research hotspot and difficulty in the field of computer vision, pedestrian detection has been widely used in intelligent driving and traffic monitoring. The popular detection method at present uses region proposal network (RPN) to generate candidate regions, and then classifies the regions. But the RPN produces many erroneous candidate areas, causing region proposals for false positives to increase. This letter uses improved residual attention network to capture the visual attention map of images, then normalized to get the attention score map. The attention score map is used to guide the RPN network to generate more precise candidate regions containing potential target objects. The region proposals, confidence scores, and features generated by the RPN are used to train a cascaded boosted forest classifier to obtain the final results. The experimental results show that our proposed approach achieves highly competitive results on the Caltech and ETH datasets.

This paper addresses the association problem of tracking closely spaced targets in group or formation. In the Labeled Multi-Bernoulli Filter (LMB), the weight of a hypothesis is directly affected by the distance between prediction and measurement. This may generate false associations when dealing with the closely spaced multiple targets. Thus we consider utilizing structure information among the group or formation. Since, the relative position relation of the targets in group or formation varies slightly within a short time, the targets are considered as nodes of a topological structure. Then the position relation among the targets is modeled as a hypergraph. The hypergraph matching method is used to resolve the association matrix. At last, with the structure prior information introduced, the new joint cost matrix is re-derived to generate hypotheses, and the filtering recursion is implemented in a Gaussian mixture way. The simulation results show that the proposed algorithm can effectively deal with group targets and is superior to the LMB filter in tracking precision and accuracy.