Latent variable models for relational data enable us to extract the co-cluster structure underlying observed relational data. The Infinite Relational Model (IRM) is a well-known relational model for discovering co-cluster structures with an unknown number of clusters. The IRM and several related models commonly assume that the link probability between two objects depends only on their cluster assignment. However, relational models based on this assumption often lead us to extract many non-informative and unexpected clusters. This is because the cluster structures underlying real-world relationships are often blurred by biases of individual objects. To overcome this problem, we propose a multi-layered framework, which extracts a clear de-blurred co-cluster structure in the presence of object biases. Then, we propose the Multi-Layered Infinite Relational Model (MLIRM) which is a special instance of the proposed framework incorporating the IRM as a co-clustering model. Furthermore, we reveal that some relational models can be regarded as special cases of the MLIRM. We derive an efficient collapsed Gibbs sampler to perform posterior inference for the MLIRM. Experiments conducted using real-world datasets have confirmed that the proposed model successfully extracts clear and interpretable cluster structures from real-world relational data.

Depth image based rendering (DIBR), which is utilized to render virtual views with a color image and the corresponding depth map, is one of the key procedures in the 2D to 3D conversion process. However, some troubling problems, such as depth edge misalignment, disocclusion occurrences and cracks at resampling, still exist in current DIBR systems. To solve these problems, in this paper, we present a robust depth image based rendering scheme for stereoscopic view synthesis. The cores of the proposed scheme are two depth map filters which share a common domain transform based filtering framework. As a first step, a filter of this framework is carried out to realize texture-depth boundary alignments and directional disocclusion reduction smoothing simultaneously. Then after depth map 3D warping, another adaptive filter is used on the warped depth maps with delivered scene gradient structures to further diminish the remaining cracks and noises. Finally, with the optimized depth map of the virtual view, backward texture warping is adopted to retrieve the final texture virtual view. The proposed scheme enables to yield visually satisfactory results for high quality 2D to 3D conversion. Experimental results demonstrate the excellent performances of the proposed approach.

In this paper, we apply extended regularized channel inversion precoding to address the multiuser multiantenna downlink transmission problem. Different from conventional regularized channel inversion precoding, extended RCI precoding considers non-homogeneous channels, adjusts more regularization parameters, and exploits the information gained by inverting the covariance matrix of the channel. Two ways of determining the regularization parameters are investigated. First, the parameters can be determined by solving a max-min SINR problem. The constraints of the problem can be transformed to the second-order cone (SOC) constraints. The optimal solution of the problem can be obtained by iteratively solving a second-order cone programming (SOCP) problem. In order to reduce the computational complexity, a one-shot algorithm is proposed. Second, the sum-rate maximization problem is discussed. The simple gradient-based method is used to solve the problem and get the regularization parameters. The simulation results indicate that the proposed algorithms exhibit improved max-min SINR performance and sum-rate performance over RCI precoding.

Fully integrated CMOS power amplifiers (PAs) with a two-winding and single-winding combined transformer (TS transformer) are presented. The general analysis of the TS transformer and other power-combining transformers, i.e., the series-combining transformer and parallel-combining transformer, is presented in terms of the transformer design parameters. Compared with other power-combining transformers, the proposed power-combining TS transformer offers high-efficiency with a compact form factor. In addition, a fully integrated CMOS PA using the TS transformer with multi-gated transistors (MGTRs) and adaptive bias circuit has been proposed to improve linearity. The proposed PAs are implemented using 65-nm CMOS technology. The implemented PA with the TS transformer achieves a saturated output power of 26.7 dBm with drain efficiency (DE) of 47.7%. The PA achieves 20.13-dBm output power with 21.4% DE while satisfying the -25-dB error vector magnitude (EVM) requirement when tested with the WLAN 802.11g signal. The implemented PA using the TS transformer with MGTRs and adaptive bias circuit achieves the -30-dB EVM requirement up to an output power of 17.13 dBm with 10.43% DE when tested using the WLAN 802.11ac signal.

In this paper, for any given prime power q, using Helleseth-Gong sequences with ideal auto-correlation property, we propose a class of perfect sequences of length (qm-1)/(q-1). As an application, a subclass of constructed perfect sequences is used to design optimal and perfect difference systems of sets.

Malware has become a growing threat as malware writers have learned that signature-based detectors can be easily evaded by packing the malware. Packing is a major challenge to malware analysis. The generic unpacking approach is the major solution to the threat of packed malware, and it is based on the intrinsic nature of the execution of packed executables. That is, the original code should be extracted in memory and get executed at run-time. The existing generic unpacking approaches need a simulated environment to monitor the executing of the packed executables. Unfortunately, the simulated environment is easily detected by the environment-sensitive packers. It makes the existing generic unpacking approaches easily evaded by the packer. In this paper, we propose a novel unpacking approach, BareUnpack, to monitor the execution of the packed executables on the bare-metal operating system, and then extracts the hidden code of the executable. BareUnpack does not need any simulated environment (debugger, emulator or VM), and it works on the bare-metal operating system directly. Our experimental results show that BareUnpack can resist the environment-sensitive packers, and improve the unpacking effectiveness, which outperforms other existing unpacking approaches.

The learning with errors (LWE) problem is considered as one of the most compelling candidates as the security base for the post-quantum cryptosystems. For the application of LWE based cryptographic schemes, the concrete parameters are necessary: the length n of secret vector, the moduli q and the deviation σ. In the middle of 2016, Germany TU Darmstadt group initiated the LWE Challenge in order to assess the hardness of LWE problems. There are several approaches to solve the LWE problem via reducing LWE to other lattice problems. Xu et al.'s group solved some LWE Challenge instances using Liu-Nguyen's adapted enumeration technique (reducing LWE to BDD problem) [23] and they published this result at ACNS 2017 [32]. In this paper, at first, we applied the progressive BKZ on the LWE challenge cases of σ/q=0.005 using Kannan's embedding technique. We can intuitively observe that the embedding technique is more efficient with the embedding factor M closer to 1. Then we will analyze the optimal number of samples m for a successful attack on LWE case with secret length of n. Thirdly based on this analysis, we show the practical cost estimations using the precise progressive BKZ simulator. Simultaneously, our experimental results show that for n ≥ 55 and the fixed σ/q=0.005, the embedding technique with progressive BKZ is more efficient than Xu et al.'s implementation of the enumeration algorithm in [32][14]. Moreover, by our parameter setting, we succeed in solving the LWE Challenge over (n,σ/q)=(70, 0.005) using 216.8 seconds (32.73 single core hours).

A tremendous amount of knowledge is present in the ever-growing scientific literature. In order to efficiently grasp such knowledge, various computational tasks are proposed that train machines to read and analyze scientific documents. One of these tasks, Scientific Relation Extraction, aims at automatically capturing scientific semantic relationships among entities in scientific documents. Conventionally, only a limited number of commonly used knowledge bases, such as Wikipedia, are used as a source of background knowledge for relation extraction. In this work, we hypothesize that unannotated scientific papers could also be utilized as a source of external background information for relation extraction. Based on our hypothesis, we propose a model that is capable of extracting background information from unannotated scientific papers. Our experiments on the RANIS corpus [1] prove the effectiveness of the proposed model on relation extraction from scientific articles.

As LSI chips integrate more transistors and the operating power supply voltage decreases, LSI chips are becoming more vulnerable to the soft error caused by neutrons induced from cosmic rays. The soft error is detected by comparing the duplicated operation results in double modular redundancy (DMR) and the error is corrected by re-executing necessary operations. In this paper, based on the error recovery scheme of re-executing necessary operations, the minimization of the vote operations for error checking with respect to given resource constraints is considered. An ILP model for the optimal solution to the problem is presented and a heuristic algorithm is proposed to minimize the vote operations.

Pseudorandom number generators are required to generate pseudorandom numbers which have good statistical properties as well as unpredictability in cryptography. An m-sequence is a linear feedback shift register sequence with maximal period over a finite field. M-sequences have good statistical properties, however we must nonlinearize m-sequences for cryptographic purposes. A geometric sequence is a sequence given by applying a nonlinear feedforward function to an m-sequence. Nogami, Tada and Uehara proposed a geometric sequence whose nonlinear feedforward function is given by the Legendre symbol, and showed the period, periodic autocorrelation and linear complexity of the sequence. Furthermore, Nogami et al. proposed a generalization of the sequence, and showed the period and periodic autocorrelation. In this paper, we first investigate linear complexity of the geometric sequences. In the case that the Chan-Games formula which describes linear complexity of geometric sequences does not hold, we show the new formula by considering the sequence of complement numbers, Hasse derivative and cyclotomic classes. Under some conditions, we can ensure that the geometric sequences have a large linear complexity from the results on linear complexity of Sidel'nikov sequences. The geometric sequences have a long period and large linear complexity under some conditions, however they do not have the balance property. In order to construct sequences that have the balance property, we propose interleaved sequences of the geometric sequence and its complement. Furthermore, we show the periodic autocorrelation and linear complexity of the proposed sequences. The proposed sequences have the balance property, and have a large linear complexity if the geometric sequences have a large one.

The response times of solid state drives (SSDs) have decreased dramatically due to the growing use of non-volatile memory express (NVMe) devices. Such devices have response times of less than 100 micro seconds on average. The response times of all-flash-array systems have also decreased dramatically through the use of NVMe SSDs. However, there are applications, particularly virtual desktop infrastructure and in-memory database systems, that require storage systems with even shorter response times. Their workloads tend to contain many input-output (IO) concentrations, which are aggregations of IO accesses. They target narrow regions of the storage volume and can continue for up to an hour. These narrow regions occupy a few percent of the logical unit number capacity, are the target of most IO accesses, and appear at unpredictable logical block addresses. To drastically reduce the response times for such workloads, we developed an automated tiered storage system called “automated tiered storage with fast memory and slow flash storage” (ATSMF) in which the data in targeted regions are migrated between storage devices depending on the predicted remaining duration of the concentration. The assumed environment is a server with non-volatile memory and directly attached SSDs, with the user applications executed on the server as this reduces the average response time. Our system predicts the effect of migration by using the previously monitored values of the increase in response time during migration and the change in response time after migration. These values are consistent for each type of workload if the system is built using both non-volatile memory and SSDs. In particular, the system predicts the remaining duration of an IO concentration, calculates the expected response-time increase during migration and the expected response-time decrease after migration, and migrates the data in the targeted regions if the sum of response-time decrease after migration exceeds the sum of response-time increase during migration. Experimental results indicate that ATSMF is at least 20% faster than flash storage only and that its memory access ratio is more than 50%.

Recently, it has been reported that malicious third-party IC vendors often insert hardware Trojans into their products. Especially in IC design step, malicious third-party vendors can easily insert hardware Trojans in their products and thus we have to detect them efficiently. In this paper, we propose a machine-learning-based hardware-Trojan detection method for gate-level netlists using multi-layer neural networks. First, we extract 11 Trojan-net feature values for each net in a netlist. After that, we classify the nets in an unknown netlist into a set of Trojan nets and that of normal nets using multi-layer neural networks. By experimentally optimizing the structure of multi-layer neural networks, we can obtain an average of 84.8% true positive rate and an average of 70.1% true negative rate while we can obtain 100% true positive rate in some of the benchmarks, which outperforms the existing methods in most of the cases.

Learning semantic representation for translation context is beneficial to statistical machine translation (SMT). Previous efforts have focused on implicitly encoding syntactic and semantic knowledge in translation context by neural networks, which are weak in capturing explicit structural syntax information. In this paper, we propose a new neural network with a tree-based convolutional architecture to explicitly learn structural syntax information in translation context, thus improving translation prediction. Specifically, we first convert parallel sentences with source parse trees into syntax-based linear sequences based on a minimum syntax subtree algorithm, and then define a tree-based convolutional network over the linear sequences to learn syntax-based context representation and translation prediction jointly. To verify the effectiveness, the proposed model is integrated into phrase-based SMT. Experiments on large-scale Chinese-to-English and German-to-English translation tasks show that the proposed approach can achieve a substantial and significant improvement over several baseline systems.

Recent studies utilize multiple kernel learning to deal with incomplete-data problem. In this study, we introduce new methods that do not only complete multiple incomplete kernel matrices simultaneously, but also allow control of the flexibility of the model by parameterizing the model matrix. By imposing restrictions on the model covariance, overfitting of the data is avoided. A limitation of kernel matrix estimations done via optimization of an objective function is that the positive definiteness of the result is not guaranteed. In view of this limitation, our proposed methods employ the LogDet divergence, which ensures the positive definiteness of the resulting inferred kernel matrix. We empirically show that our proposed restricted covariance models, employed with LogDet divergence, yield significant improvements in the generalization performance of previous completion methods.

We analyze the behaviors of the FXLMS algorithm using a statistical-mechanical method. The cross-correlation between a primary path and an adaptive filter and the autocorrelation of the adaptive filter are treated as macroscopic variables. We obtain simultaneous differential equations that describe the dynamical behaviors of the macroscopic variables under the condition that the tapped-delay line is sufficiently long. The obtained equations are deterministic and closed-form. We analytically solve the equations to obtain the correlations and finally compute the mean-square error. The obtained theory can quantitatively predict the behaviors of computer simulations including the cases of both not only white but also nonwhite reference signals. The theory also gives the upper limit of the step size in the FXLMS algorithm.

We normally hold a lot of confidential information in hard disk drives and solid-state drives. When we want to erase such information to prevent the leakage, we have to overwrite the sequence of information with a sequence of symbols independent of the information. The overwriting is needed only at places where overwritten symbols are different from original symbols. Then, the cost of overwrites such as the number of overwritten symbols to erase information is important. In this paper, we clarify the minimum cost such as the minimum number of overwrites to erase information under weak and strong independence criteria. The former (resp. the latter) criterion represents that the mutual information between the original sequence and the overwritten sequence normalized (resp. not normalized) by the length of the sequences is less than a given desired value.

In this paper, we investigate how to obtain binary locally repairable codes (LRCs) with good locality and availability from binary Simplex codes. We first propose a Combination code having the generator matrix with all the columns of positive weights less than or equal to a given value. Such a code can be also obtained by puncturing all the columns of weights larger than a given value from a binary Simplex Code. We call by block-puncturing such puncturing method. Furthermore, we suggest a heuristic puncturing method, called subblock-puncturing, that punctures a few more columns of the largest weight from the Combination code. We determine the minimum distance, locality, availability, joint information locality, joint information availability of Combination codes in closed-form. We also demonstrate the optimality of the proposed codes with certain choices of parameters in terms of some well-known bounds.

As energy efficiency has become a major design constraint or objective, heterogeneous manycore architectures have emerged as mainstream target platforms not only in server systems but also in embedded systems. Manycore accelerators such as GPUs are getting also popular in embedded domains, as well as the heterogeneous CPU cores. However, as the number of cores in an embedded GPU is far less than that of a server GPU, it is important to utilize both heterogeneous multi-core CPUs and GPUs to achieve the desired throughput with the minimal energy consumption. In this paper, we present a case study of mapping LBP-based face detection onto a recent CPU-GPU heterogeneous embedded platform, which exploits both task parallelism and data parallelism to achieve maximal energy efficiency with a real-time constraint. We first present the parallelization technique of each task for the GPU execution, then we propose performance and energy models for both task-parallel and data-parallel executions on heterogeneous processors, which are used in design space exploration for the optimal mapping. The design space is huge since not only processor heterogeneity such as CPU-GPU and big.LITTLE, but also various data partitioning ratios for the data-parallel execution on these heterogeneous processors are considered. In our case study of LBP face detection on Exynos 5422, the estimation error of the proposed performance and energy models were on average -2.19% and -3.67% respectively. By systematically finding the optimal mappings with the proposed models, we could achieve 28.6% less energy consumption compared to the manual mapping, while still meeting the real-time constraint.

Crack detection is a vital task to maintain a bridge's health and safety condition. Traditional computer-vision based methods easily suffer from disturbance of noise and clutters for a real bridge inspection. To address this limitation, we propose a two-stage crack detection approach based on Convolutional Neural Networks (CNN) in this letter. A predictor of small receptive field is exploited in the first detection stage, while another predictor of large receptive field is used to refine the detection results in the second stage. Benefiting from data fusion of confidence maps produced by both predictors, our method can predict the probability belongs to cracked areas of each pixel accurately. Experimental results show that the proposed method is superior to an up-to-date method on real concrete surface images.

In the past two decades, many generalized cyclotomic sequences have been constructed and they have been used in cryptography and communication systems for their high linear complexity and low autocorrelation. But there are a few of papers focusing on the 2-adic complexities of such sequences. In this paper, we first give a property of a class of Gaussian periods based on Whiteman's generalized cyclotomic classes of order 4. Then, as an application of this property, we study the 2-adic complexity of a class of Whiteman's generalized cyclotomic sequences constructed from two distinct primes p and q. We prove that the 2-adic complexity of this class of sequences of period pq is lower bounded by pq-p-q-1. This lower bound is at least greater than one half of its period and thus it shows that this class of sequences can resist against the rational approximation algorithm (RAA) attack.