Ultra-wideband (UWB) technology has attracted much attention recently due to its high data rate and low emission power. Its media access control (MAC) protocol, WiMedia MAC, promises a lot of facilities for high-speed and high-quality wireless communication. However, these benefits in turn involve a large amount of computational load, which challenges the traditional uniprocessor architecture based implementation method to provide the required performance. However, the constrained cost and power budget, on the other hand, makes using commercial multiprocessor solutions unrealistic. In this paper, a low-cost and energy-efficient multiprocessor system-on-chip (MPSoC), which tackles at once the aspects of system design, software migration and hardware architecture, is presented for the implementation of UWB MAC layer. Experimental results show that the proposed MPSoC, based on four simple RISC processors and shared-memory infrastructure, achieves up to 45% performance improvement and 65% power saving, but takes 15% less area than the uniprocessor implementation.

抄録全体を表示

This paper proposes a novel dynamic Scratch-pad Memory allocation strategy to optimize the energy consumption of the memory sub-system. Firstly, the whole program execution process is sliced into several time slots according to the temporal dimension; thereafter, a Time-Slotted Cache Conflict Graph (TSCCG) is introduced to model the behavior of Data Cache (D-Cache) conflicts within each time slot. Then, Integer Nonlinear Programming (INP) is implemented, which can avoid time-consuming linearization process, to select the most profitable data pages. Virtual Memory System (VMS) is adopted to remap those data pages, which will cause severe Cache conflicts within a time slot, to SPM. In order to minimize the swapping overhead of dynamic SPM allocation, a novel SPM controller with a tightly coupled DMA is introduced to issue the swapping operations without CPU's intervention. Last but not the least, this paper discusses the fluctuation of system energy profit based on different MMU page size as well as the Time Slot duration quantitatively. According to our design space exploration, the proposed method can optimize all of the data segments, including global data, heap and stack data in general, and reduce the total energy consumption by 27.28% on average, up to 55.22% with a marginal performance promotion. And comparing to the conventional static CCG (Cache Conflicts Graph), our approach can obtain 24.7% energy profit on average, up to 30.5% with a sight boost in performance.

抄録全体を表示

Distributed applications and services have become pervasive in our society due to the widespread use of internet and mobile devices. There are urgent demands to efficiently ensure safety and correctness of such software. A session-type system is a framework to statically check whether communication descriptions conform to certain protocols. They are shown to be effective yet simple enough to fit in harmony with existing programming languages. In the original session type system, the subject reduction property does not hold. This paper establishes a conservative extension of the original session type system with the subject reduction property. Finally, it is also shown that our typing rule properly extends the set of typeable processes.

抄録全体を表示

The objective of pool-based incremental active learning is to choose a sample to label from a pool of unlabeled samples in an incremental manner so that the generalization error is minimized. In this scenario, the generalization error often hits a minimum in the middle of the incremental active learning procedure and then it starts to increase. In this paper, we address the problem of early labeling stopping in probabilistic classification for minimizing the generalization error and the labeling cost. Among several possible strategies, we propose to stop labeling when the empirical class-posterior approximation error is maximized. Experiments on benchmark datasets demonstrate the usefulness of the proposed strategy.

抄録全体を表示

This paper proposes a new generative model which can deal with rotational data variations by extending Separable Lattice 2-D HMMs (SL2D-HMMs). In image recognition, geometrical variations such as size, location and rotation degrade the performance. Therefore, the appropriate normalization processes for such variations are required. SL2D-HMMs can perform an elastic matching in both horizontal and vertical directions; this makes it possible to model invariance to size and location. To deal with rotational variations, we introduce additional HMM states which represent the shifts of the state alignments among the observation lines in a particular direction. Face recognition experiments show that the proposed method improves the performance significantly for rotational variation data.

抄録全体を表示

The performance of English automatic speech recognition systems decreases when recognizing spontaneous speech mainly due to multiple pronunciation variants in the utterances. Previous approaches address this problem by modeling the alteration of the pronunciation on a phoneme to phoneme level. However, the phonetic transformation effects induced by the pronunciation of the whole sentence have not yet been considered. In this article, the sequence-based pronunciation variation is modeled using a noisy channel approach where the spontaneous phoneme sequence is considered as a “noisy” string and the goal is to recover the “clean” string of the word sequence. Hereby, the whole word sequence and its effect on the alternation of the phonemes will be taken into consideration. Moreover, the system not only learns the phoneme transformation but also the mapping from the phoneme to the word directly. In this study, first the phonemes will be recognized with the present recognition system and afterwards the pronunciation variation model based on the noisy channel approach will map from the phoneme to the word level. Two well-known natural language processing approaches are adopted and derived from the noisy channel model theory: Joint-sequence models and statistical machine translation. Both of them are applied and various experiments are conducted using microphone and telephone of spontaneous speech.

抄録全体を表示

In this paper, we propose Hidden Conditional Neural Fields (HCNF) for continuous phoneme speech recognition, which are a combination of Hidden Conditional Random Fields (HCRF) and a Multi-Layer Perceptron (MLP), and inherit their merits, namely, the discriminative property for sequences from HCRF and the ability to extract non-linear features from an MLP. HCNF can incorporate many types of features from which non-linear features can be extracted, and is trained by sequential criteria. We first present the formulation of HCNF and then examine three methods to further improve automatic speech recognition using HCNF, which is an objective function that explicitly considers training errors, provides a hierarchical tandem-style feature and includes a deep non-linear feature extractor for the observation function. We show that HCNF can be trained realistically without any initial model and outperforms HCRF and the triphone hidden Markov model trained by the minimum phone error (MPE) manner using experimental results for continuous English phoneme recognition on the TIMIT core test set and Japanese phoneme recognition on the IPA 100 test set.

抄録全体を表示

Template matching for image sequences captured with a moving camera is very important for several applications such as Robot Vision, SLAM, ITS, and video surveillance systems. However, it is difficult to realize accurate template matching using only visual feature information such as HSV histograms, edge histograms, HOG histograms, and SIFT features, because it is affected by several phenomena such as illumination change, viewpoint change, size change, and noise. In order to realize robust tracking, structure information such as the relative position of each part of the object should be considered. In this paper, we propose a method that considers both visual feature information and structure information. Experiments show that the proposed method realizes robust tracking and determine the relationships between object parts in the scenes and those in the template.

抄録全体を表示

Two types of similarities between words have been studied in the natural language processing community: synonymy and relational similarity. A high degree of similarity exist between synonymous words. On the other hand, a high degree of relational similarity exists between analogous word pairs. We present and empirically test a hypothesis that links these two types of similarities. Specifically, we propose a method to measure the degree of synonymy between two words using relational similarity between word pairs as a proxy. Given two words, first, we represent the semantic relations that hold between those words using lexical patterns. We use a sequential pattern clustering algorithm to identify different lexical patterns that represent the same semantic relation. Second, we compute the degree of synonymy between two words using an inter-cluster covariance matrix. We compare the proposed method for measuring the degree of synonymy against previously proposed methods on the Miller-Charles dataset and the WordSimilarity-353 dataset. Our proposed method outperforms all existing Web-based similarity measures, achieving a statistically significant Pearson correlation coefficient of 0.867 on the Miller-Charles dataset.

抄録全体を表示

We investigate a sparsely encoded Hopfield model with unit replacement by using a statistical mechanical method called self-consistent signal-to-noise analysis. We theoretically obtain a relation between the storage capacity and the number of replacement units for each sparseness a. Moreover, we compare the unit replacement model with the forgetting model in terms of the network storage capacity. The results show that the unit replacement model has a finite value of the optimal sparseness on an open interval 0 (1/2 coding) <a<1 (the limit of sparseness) to maximize the storage capacity for a large number of replacement units, although the forgetting model does not.

抄録全体を表示

This paper presents a new hybrid speed function needed to perform image segmentation within the level-set framework. The proposed speed function uses both the boundary and region information of objects to achieve robust and accurate segmentation results. This speed function provides a general form that incorporates the robust alignment term as a part of the driving force for the proper edge direction of an active contour, an active region term derived from the region partition scheme, and the smoothing term for regularization. First, we use an external force for active contours as the Gradient Vector Flow field. This is computed as the diffusion of gradient vectors of a gray level edge map derived from an image. Second, we partition the image domain by progressively fitting statistical models to the intensity of each region. Here we adopt two Gaussian distributions to model the intensity distribution of the inside and outside of the evolving curve partitioning the image domain. Third, we use the active contour model that has the computation of geodesics or minimal distance curves, which allows stable boundary detection when the model's gradients suffer from large variations including gaps or noise. Finally, we test the accuracy and robustness of the proposed method for various medical images. Experimental results show that our method can properly segment low contrast, complex images.

抄録全体を表示

This paper proposes a new first-scan method for two-scan labeling algorithms. In the first scan, our proposed method first scans every fourth image line, and processes the scan line and its two neighbor lines. Then, it processes the remaining lines from top to bottom one by one. Our method decreases the average number of times that must be checked to process a foreground pixel will; thus, the efficiency of labeling can be improved.

抄録全体を表示

The time efficiency of aerial video stitching is still an open problem due to the huge amount of input frames, which usually results in prohibitive complexities in both image registration and blending. In this paper, we propose an efficient framework aiming to stitch aerial videos in real time. Reasonable distortions are allowed as a tradeoff for acceleration. Instead of searching for globally optimized solutions, we directly refine frame positions with sensor data to compensate for the accumulative error in alignment. A priority scan method is proposed to select pixels within overlapping area into the final panorama for blending, which avoids complicated operations like weighting or averaging on pixels. Experiments show that our method can generate satisfying results at very competitive speed.

抄録全体を表示

This letter proposes a new image colorization algorithm based on the sparse optimization. Introducing some assumptions, a problem of recovering a color image from a grayscale image with the small number of known color pixels is formulated as a mixed l₀/l₁ norm minimization, and an iterative reweighted least squares (IRLS) algorithm is proposed. Numerical examples show that the proposed algorithm colorizes the grayscale image efficiently.

抄録全体を表示

Person re-identification is challenging due to illumination changes and viewpoint variations in the multi-camera environment. In this paper, we propose a novel spatial pyramid color representation (SPCR) and a local region matching scheme, to explore person appearance for re-identification. SPCR effectively integrates color layout into histogram, forming an informative global feature. Local region matching utilizes region statistics, which is described by covariance feature, to find appearance correspondence locally. Our approach shows robustness to illumination changes and slight viewpoint variations. Experiments on a public dataset demonstrate the performance superiority of our proposal over state-of-the-art methods.

抄録全体を表示

A new training algorithm for the chaotic Adachi Neural Network (AdNN) is investigated. The classical training algorithm for the AdNN and it's variants is usually a “one-shot” learning, for example, the Outer Product Rule (OPR) is the most used. Although the OPR is effective for conventional neural networks, its effectiveness and adequateness for Chaotic Neural Networks (CNNs) have not been discussed formally. As a complementary and tentative work in this field, we modified the AdNN's weights by enforcing an unsupervised Hebbian rule. Experimental analysis shows that the new weighted AdNN yields even stronger dynamical associative memory and pattern recognition phenomena for different settings than the primitive AdNN.

抄録全体を表示