In this paper, we propose a new method for the bias-dependent parameter extraction of a MOSFET, which covers DC to over 100 GHz. The DC MOSFET model provided by the chip foundry is assumed to be correct, and the core DC characteristics are designed to be asymptotically recovered at low frequencies. This is carried out by representing the corrections required at high frequencies using a bias-dependent Y matrix, assuming that a parasitic nonlinear two-port matrix (Y-wrapper) is connected in parallel with the core MOSFET. The Y-wrapper can also handle the nonreciprocity of the parasitic components, that is, the asymmetry of the Y matrix. The reliability of the Y-wrapper model is confirmed through the simulation and measurement of a one-stage common-source amplifier operating at several bias points. This paper will not discuss about non-linearity.

This paper presents fundamental concepts, overall process and recent research issues of Mining Software Repositories. The data sources such as source control systems, bug tracking systems or archived communications, data types and techniques used for general MSR problems are also presented. Finally, evaluation approaches, opportunities and challenge issues are given.

This paper proposes a video-quality estimation method based on a reduced-reference model for realtime quality monitoring in video streaming services. The proposed method chooses representative-luminance values for individual original-video frames at a server side and transmits those values, along with the pixel-position information of the representative-luminance values in each frame. On the basis of this information, peak signal-to-noise ratio (PSNR) values at client sides can be estimated. This enables realtime monitoring of video-quality degradation by transmission errors. Experimental results show that accurate PSNR estimation can be achieved with additional information at a low bit rate. For SDTV video sequences which are encoded at 1 to 5 Mbps, accurate PSNR estimation (correlation coefficient of 0.92 to 0.95) is achieved with small amount of additional information of 10 to 50 kbps. This enables accurate realtime quality monitoring in video streaming services without average video-quality degradation.

In this paper, a novel framework for extracting visual feature-based keyword relationships from an image database is proposed. From the characteristic that a set of relevant keywords tends to have common visual features, the keyword relationships in a target image database are extracted by using the following two steps. First, the relationship between each keyword and its corresponding visual features is modeled by using a classifier. This step enables detection of visual features related to each keyword. In the second step, the keyword relationships are extracted from the obtained results. Specifically, in order to measure the relevance between two keywords, the proposed method removes visual features related to one keyword from training images and monitors the performance of the classifier obtained for the other keyword. This measurement is the biggest difference from other conventional methods that focus on only keyword co-occurrences or visual similarities. Results of experiments conducted using an image database showed the effectiveness of the proposed method.

The development of ontology at the instance level requires the extraction of the terms defining the instances from various data sources. These instances then are linked to the concepts of the ontology, and relationships are created between these instances for the next step. However, before establishing links among data, ontology engineers must classify terms or instances from a web document into an ontology concept. The tool for help ontology engineer in this task is called ontology population. The present research is not suitable for ontology development applications, such as long time processing or analyzing large or noisy data sets. OntoPop system introduces a methodology to solve these problems, which comprises two parts. First, we select meaningful features from syntactic relations, which can produce more significant features than any other method. Second, we differentiate feature meaning and reduce noise based on latent semantic analysis. Experimental evaluation demonstrates that the OntoPop works well, significantly out-performing the accuracy of 49.64%, a learning accuracy of 76.93%, and executes time of 5.46 second/instance.

We describe a cell-based connected component labeling algorithm to calculate the 0th and 1st moment features as the attributes for labeled regions. These can be used to indicate their sizes and positions for multi-object extraction. Based on the additivity in moment features, the cell-based labeling algorithm can label divided cells of a certain size in an image by scanning the image only once to obtain the moment features of the labeled regions with remarkably reduced computational complexity and memory consumption for labeling. Our algorithm is a simple-one-time-scan cell-based labeling algorithm, which is suitable for hardware and parallel implementation. We also compared it with conventional labeling algorithms. The experimental results showed that our algorithm is faster than conventional raster-scan labeling algorithms.

This letter realizes direction of arrival (DOA) estimation by exploiting the noise subspace based estimator. Since single subspace feature extraction fails to achieve satisfactory results under unknown noise fields, we propose a two-step subspace feature extraction technique that is effective even in these fields. When a new noise subspace is attained, the proposed estimator without prewhitening can form the maximizing orthogonality especially for unknown noise fields. Simulation results confirm the effectiveness of the proposed technique.

In the past few years, discriminant analysis and manifold learning have been widely used in feature extraction. Recently, the sparse representation technique has advanced the development of pattern recognition. In this paper, we combine both discriminant analysis and manifold learning with sparse representation technique and propose a novel feature extraction approach named sparsity preserving embedding with manifold learning and discriminant analysis. It seeks an embedded space, where not only the sparse reconstructive relations among original samples are preserved, but also the manifold and discriminant information of both original sample set and the corresponding reconstructed sample set is maintained. Experimental results on the public AR and FERET face databases show that our approach outperforms relevant methods in recognition performance.

Modern web applications incorporate many programmatic frameworks and APIs that are often pushed to the client-side with most of the application logic while contents are the result of mashing up several resources from different origins. Such applications are threatened by attackers that often attempts to inject directly, or by leveraging a stepstone website, script codes that perform malicious operations. Web scripting based malware proliferation is being more and more industrialized with the drawbacks and advantages that characterize such approach: on one hand, we are witnessing a lot of samples that exhibit the same characteristics which make these easy to detect, while on the other hand, professional developers are continuously developing new attack techniques. While obfuscation is still a debated issue within the community, it becomes clear that, with new schemes being designed, this issue cannot be ignored anymore. Because many proposed countermeasures confess that they perform better on unobfuscated contents, we propose a 2-stage technique that first relieve the burden of obfuscation by emulating the deobfuscation stage before performing a static abstraction of the analyzed sample's functionalities in order to reveal its intent. We support our proposal with evidence from applying our technique to real-life examples and provide discussion on performance in terms of time, as well as possible other applications of proposed techniques in the areas of web crawling and script classification. Additionally, we claim that such approach can be generalized to other scripting languages similar to JavaScript.

We propose a novel background and foreground estimation algorithm in MAP-MRF approach for binarization of degraded document image. In the proposed algorithm, an assumption that background whiteness and foreground blackness is not employed differently from the conventional algorithm, and we employ character's irregularities based on local statistics. This makes the method possible to apply to the image with various colored characters, ex. outlined characters by colored background. The effectiveness and the validity are shown by applying the proposed method to various degraded document images.

In this paper, we present an approach of detecting speech presence for which the decision rule is based on a combination of multiple features using a sigmoid function. A minimum classification error (MCE) training is used to update the weights adjustment for the combination. The features, consisting of three parameters: the ratio of ZCR, the spectral energy, and spectral entropy, are combined linearly with weights derived from the sub-band domain. First, the Bark-scale wavelet decomposition (BSWD) is used to split the input speech into 24 critical sub-bands. Next, the feature parameters are derived from the selected frequency sub-band to form robust voice feature parameters. In order to discard the seriously corrupted frequency sub-band, a strategy of adaptive frequency sub-band extraction (AFSE) dependant on the sub-band SNR is then applied to only the frequency sub-band used. Finally, these three feature parameters, which only consider the useful sub-band, are combined through a sigmoid type function incorporating optimal weights based on MSE training to detect either a speech present frame or a speech absent frame. Experimental results show that the performance of the proposed algorithm is superior to the standard methods such as G.729B and AMR2.

In this paper, an adaptive block-based text line extraction algorithm is proposed. Three global and two local parameters are defined to adapt the method to various handwritings in different languages. A document image is segmented into several overlapping blocks. The skew of each block is estimated. Text block is de-skewed by using the estimated skew angle. Text regions are detected in the de-skewed text block. A number of data points are extracted from the detected text regions in each block. These data points are used to estimate the paths of text lines. By thinning the background of the image including text line paths, text line boundaries or separators are estimated. Furthermore, an algorithm is proposed to assign to the extracted text lines the connected components which have intersections with the estimated separators. Extensive experiments on different standard datasets in various languages demonstrate that the proposed algorithm outperforms previous methods.

Based on sliding-window rule application and extraction filtering, we present a framework for extracting multi-slot frames describing chemical reactions from Thai free text with unknown target-phrase boundaries. A supervised rule learning algorithm is employed for automatic construction of pattern-based extraction rules from hand-tagged training phrases. A filtering method is devised for removal of incorrect extraction results based on features observed from text portions appearing between adjacent slot fillers in source documents. Extracted reaction frames are represented as concept expressions in description logics and are used as metadata for document indexing. A document knowledge base supporting semantics-based information retrieval is constructed by integrating document metadata with domain-specific ontologies.

Due to the limitations of language-processing tools for the Thai language, pattern-based information extraction from Thai documents requires supplementary techniques. Based on sliding-window rule application and extraction filtering, we present a framework for extracting semantic information from medical-symptom phrases with unknown boundaries in Thai unstructured-text information entries. A supervised rule learning algorithm is employed for automatic construction of information extraction rules from hand-tagged training symptom phrases. Two filtering components are introduced: one uses a classification model to predict rule application across a symptom-phrase boundary based on instantiation features of rule internal wildcards, the other uses weighted classification confidence to resolve conflicts arising from overlapping extractions. In our experimental study, we focus our attention on two basic types of symptom phrasal descriptions: one is concerned with abnormal characteristics of some observable entities and the other with human-body locations at which primitive symptoms appear. The experimental results show that the filtering components improve precision while preserving recall satisfactorily.

We propose a blind dereverberation method based on spectral subtraction using a multi-channel least mean squares (MCLMS) algorithm for distant-talking speech recognition. In a distant-talking environment, the channel impulse response is longer than the short-term spectral analysis window. By treating the late reverberation as additive noise, a noise reduction technique based on spectral subtraction was proposed to estimate the power spectrum of the clean speech using power spectra of the distorted speech and the unknown impulse responses. To estimate the power spectra of the impulse responses, a variable step-size unconstrained MCLMS (VSS-UMCLMS) algorithm for identifying the impulse responses in a time domain is extended to a frequency domain. To reduce the effect of the estimation error of the channel impulse response, we normalize the early reverberation by cepstral mean normalization (CMN) instead of spectral subtraction using the estimated impulse response. Furthermore, our proposed method is combined with conventional delay-and-sum beamforming. We conducted recognition experiments on a distorted speech signal simulated by convolving multi-channel impulse responses with clean speech. The proposed method achieved a relative error reduction rate of 22.4% in relation to conventional CMN. By combining the proposed method with beamforming, a relative error reduction rate of 24.5% in relation to the conventional CMN with beamforming was achieved using only an isolated word (with duration of about 0.6 s) to estimate the spectrum of the impulse response.

This paper proposes a novel image segmentation method based on Clausius entropy and adaptive Gaussian mixture model for detecting moving objects in a complex environment. The results suggest that the proposed method performs better than existing methods in extracting the foreground in various video sequences composed of multiple objects, lighting reflections, and background clutter.

A sanitizable signature scheme allows a semi-trusted party, designated by a signer, to modify pre-determined parts of a signed message without interacting with the original signer. To date, many sanitizable signature schemes have been proposed based on various cryptographic techniques. However, previous works are usually built upon the paradigm of dividing a message into submessages and applying a cryptographic primitive to each submessage. This methodology entails the computation time (and often signature length) in linear proportion to the number of sanitizable submessages. We present a new approach to constructing sanitizable signatures with constant overhead for signing and verification, irrespective of the number of submessages, both in computational cost and in signature size.

This paper proposes a synthesis method of 2-channel IIR paraunitary filter banks by successive extraction of 2-port lattice sections. When a power symmetry transfer function is given, a filter bank is realized as cascade of paraunitary 2-port lattice sections. The method can synthesize both odd- and even-order filters with Butterworth or elliptic characteristics. The number of multiplications per second can also be reduced.

This paper describes a nonlinear filter that can extract the image feature from noise corrupted image labeled self-quotient ε-filter (SQEF). SQEF is an improved self-quotient filter (SQF) to extract the image feature from noise corrupted image. Although SQF is a simple approach for feature extraction from the images, it is difficult to extract the feature when the image includes noise. On the other hand, SQEF can extract the image feature not only from clear images but also from noise corrupted images with uniform noise, Gaussian noise and impulse noise. We show the algorithm of SQEF and describe its feature when it is applied to uniform noise corrupted image, Gaussian noise corrupted image and impulse noise corrupted image. Experimental results are also shown to confirm the effectiveness of the proposed method.

We propose a new parameter-extraction approach based on a mixed-mode genetic algorithm (GA), including the efficient search-space separation and local-minima-convergence prevention process. The technique, substantially extended from our previous work, allows the designed figures-of-merit, such as internal quantum efficiency (ηi) as well as transparency current density (Jtr) of lasers and minimum noise figure (NFmin) as well as associated available gain (GA,assoc) of low-noise amplifiers (LNAs), extracted by an analytical equation-based methodology combined with an evolutionary numerical tool. Extraction results, which agree well with actually measured data, for both state-of-the-art InGaAs quantum-well lasers and advanced SiGe LNAs are presented for the first time to demonstrate this multi-parameter analysis and high-accuracy optimization.