We exhibit a simple procedure to find how classical signals should be processed in cluster-state quantum computation. Using stabilizers characterizing a cluster state, we can easily find a precise classical signal-flow that is required in performing cluster-state computation.
Sheng LI Xiao-Yuan JING Lu-Sha BIAN Shi-Qiang GAO Qian LIU Yong-Fang YAO
In this letter, a statistical uncorrelated near class discriminant (SUNCD) approach is proposed for face recognition. The optimal discriminant vector obtained by this approach can differentiate one class and its near classes, i.e., its nearest neighbor classes, by constructing the specific between-class and within-class scatter matrices and using the Fisher criterion. In this manner, SUNCD acquires all discriminant vectors class by class. Furthermore, SUNCD makes every discriminant vector satisfy locally statistical uncorrelated constraints by using the corresponding class and part of its most neighboring classes. Experiments on the public AR face database demonstrate that the proposed approach outperforms several representative discriminant methods.
Keiki TAKADAMA Kazuyuki HIROSE Hiroyasu MATSUSHIMA Kiyohiko HATTORI Nobuo NAKAJIMA
This paper proposes the sleep stage estimation method that can provide an accurate estimation for each person without connecting any devices to human's body. In particular, our method learns the appropriate multiple band-pass filters to extract the specific wave pattern of heartbeat, which is required to estimate the sleep stage. For an accurate estimation, this paper employs Learning Classifier System (LCS) as the data-mining techniques and extends it to estimate the sleep stage. Extensive experiments on five subjects in mixed health confirm the following implications: (1) the proposed method can provide more accurate sleep stage estimation than the conventional method, and (2) the sleep stage estimation calculated by the proposed method is robust regardless of the physical condition of the subject.
Ji-Soo KEUM Hyon-Soo LEE Masafumi HAGIWARA
In this letter, we propose an improved speech/ nonspeech classification method to effectively classify a multimedia source. To improve performance, we introduce a feature based on spectral duration analysis, and combine recently proposed features such as high zero crossing rate ratio (HZCRR), low short time energy ratio (LSTER), and pitch ratio (PR). According to the results of our experiments on speech, music, and environmental sounds, the proposed method obtained high classification results when compared with conventional approaches.
Pedro PERIS-LOPEZ Tieyan LI Julio C. HERNANDEZ-CASTRO
In 2006 EPCglobal and the International Organization for Standards (ISO) ratified the EPC Class-1 Generation-2 (Gen-2) and the ISO 18000-6C standards , respectively. These efforts represented major advancements in the direction of universal standardization for low-cost RFID tags. However, a cause for concern is that security issues do not seem to be properly addressed. In this paper, we propose a new lightweight RFID tag-reader mutual authentication scheme for use under the EPCglobal framework. The scheme is based on previous work by Konidala and Kim . We attempt to mitigate the weaknesses observed in the original scheme and, at the same time, consider other possible adversarial threats as well as constraints on low-cost RFID tags requirements.
In this study, a discriminative weight training is applied to a support vector machine (SVM) based speech/music classification for a 3GPP2 selectable mode vocoder (SMV). In the proposed approach, the speech/music decision rule is derived by the SVM by incorporating optimally weighted features derived from the SMV based on a minimum classification error (MCE) method. This method differs from that of the previous work in that different weights are assigned to each feature of the SMV a novel process. According to the experimental results, the proposed approach is effective for speech/music classification using the SVM.
Wen-An TSOU Wen-Shen WUEN Kuei-Ann WEN
A circuit technique to correct Vdd/PM distortion and improve efficiency as supply modulation of cascode class-E PAs has been proposed. The experimental result shows that the phase distortion can be improved from 20 degrees to 5 degrees. Moreover, a system co-simulation result demonstrated that the EVM can be improved from -17 dB to -19 dB.
Bong-Jin LEE Chi-Sang JUNG Jeung-Yoon CHOI Hong-Goo KANG
This letter describes the importance of transition regions, e.g. at phoneme boundaries, for automatic speaker recognition compared with using steady-state regions. Experimental results of automatic speaker identification tasks confirm that transition regions include the most speaker distinctive features. A possible reason for obtaining such results is described in view of articulation, in particular, the degree of freedom of articulators. These results are expected to provide useful information in designing an efficient automatic speaker recognition system.
This paper presents an extended Relief-F algorithm for nominal attribute estimation, for application to small-document classification. Relief algorithms are general and successful instance-based feature-filtering algorithms for data classification and regression. Many improved Relief algorithms have been introduced as solutions to problems of redundancy and irrelevant noisy features and to the limitations of the algorithms for multiclass datasets. However, these algorithms have only rarely been applied to text classification, because the numerous features in multiclass datasets lead to great time complexity. Therefore, in considering their application to text feature filtering and classification, we presented an extended Relief-F algorithm for numerical attribute estimation (E-Relief-F) in 2007. However, we found limitations and some problems with it. Therefore, in this paper, we introduce additional problems with Relief algorithms for text feature filtering, including the negative influence of computation similarities and weights caused by a small number of features in an instance, the absence of nearest hits and misses for some instances, and great time complexity. We then suggest a new extended Relief-F algorithm for nominal attribute estimation (E-Relief-Fd) to solve these problems, and we apply it to small text-document classification. We used the algorithm in experiments to estimate feature quality for various datasets, its application to classification, and its performance in comparison with existing Relief algorithms. The experimental results show that the new E-Relief-Fd algorithm offers better performance than previous Relief algorithms, including E-Relief-F.
Andrew FINCH Eiichiro SUMITA Satoshi NAKAMURA
This paper presents a technique for class-dependent decoding for statistical machine translation (SMT). The approach differs from previous methods of class-dependent translation in that the class-dependent forms of all models are integrated directly into the decoding process. We employ probabilistic mixture weights between models that can change dynamically on a sentence-by-sentence basis depending on the characteristics of the source sentence. The effectiveness of this approach is demonstrated by evaluating its performance on travel conversation data. We used this approach to tackle the translation of questions and declarative sentences using class-dependent models. To achieve this, our system integrated two sets of models specifically built to deal with sentences that fall into one of two classes of dialog sentence: questions and declarations, with a third set of models built with all of the data to handle the general case. The technique was thoroughly evaluated on data from 16 language pairs using 6 machine translation evaluation metrics. We found the results were corpus-dependent, but in most cases our system was able to improve translation performance, and for some languages the improvements were substantial.
Rachel Mabanag CHONG Toshihisa TANAKA
A new algorithm for simultaneously detecting and identifying invariant blurs is proposed. This is mainly based on the behavior of extrema values in an image. It is computationally simple and fast thereby making it suitable for preprocessing especially in practical imaging applications. Benefits of employing this method includes the elimination of unnecessary processes since unblurred images will be separated from the blurred ones which require deconvolution. Additionally, it can improve reconstruction performance by proper identification of blur type so that a more effective blur specific deconvolution algorithm can be applied. Experimental results on natural images and its synthetically blurred versions show the characteristics and validity of the proposed method. Furthermore, it can be observed that feature selection makes the method more efficient and effective.
Akara SOPHARAK Bunyarit UYYANONVARA Sarah BARMAN Thomas WILLIAMSON
To prevent blindness from diabetic retinopathy, periodic screening and early diagnosis are neccessary. Due to lack of expert ophthalmologists in rural area, automated early exudate (one of visible sign of diabetic retinopathy) detection could help to reduce the number of blindness in diabetic patients. Traditional automatic exudate detection methods are based on specific parameter configuration, while the machine learning approaches which seems more flexible may be computationally high cost. A comparative analysis of traditional and machine learning of exudates detection, namely, mathematical morphology, fuzzy c-means clustering, naive Bayesian classifier, Support Vector Machine and Nearest Neighbor classifier are presented. Detected exudates are validated with expert ophthalmologists' hand-drawn ground-truths. The sensitivity, specificity, precision, accuracy and time complexity of each method are also compared.
Packet classification categorizes incoming packets into multiple forwarding classes based on pre-defined filters. This categorization makes information accessible for quality of service or security handling in the network. In this paper, we propose a scheme which combines the Aggregate Bit Vector algorithm and the Pruned Tuple Space Search algorithm to improve the performance of packet classification in terms of speed and storage. We also present the procedures of incremental update. Our scheme is evaluated with filter databases of varying sizes and characteristics. The experimental results demonstrate that our scheme is feasible and scalable.
Network intrusion detection systems rely on a signature-based detection engine. When under attack or during heavy traffic, the detection engines need to make a fast decision whether a packet or a sequence of packets is normal or malicious. However, if packets have a heavy payload or the system has a great deal of attack patterns, the high cost of payload inspection severely diminishes detection performance. Therefore, it would be better to avoid unnecessary payload scans by checking the protocol fields in the packet header, before executing their heavy operations of payload inspection. When payload inspection is necessary, it is better to compare a minimum number of attack patterns. In this paper, we propose new methods to classify attack signatures and make pre-computed multi-pattern groups. Based on IDS rule analysis, we grouped the signatures of attack rules by a multi-dimensional classification method adapted to a simplified address flow. The proposed methods reduce unnecessary payload scans and make light pattern groups to be checked. While performance improvements are dependent on a given networking environment, the experimental results with the DARPA data set and university traffic show that the proposed methods outperform the most recent Snort by up to 33%.
In conjunction with a first-order Taylor series approximation of the spatial scanning vector, this letter presents an iterative multiple signal classification (MUSIC) direction-of-arrival (DOA) estimation for code-division multiple access signals. This approach leads to a simple one-dimensional optimization problem to find each iterative optimal search grid. It can not only accurately estimate DOA, but also speed up the estimating process. Computer results demonstrate the effectiveness of the proposed algorithm.
In this Letter, a robust system identification method is proposed for the generalized sidelobe canceller using dual microphones. The conventional transfer-function generalized sidelobe canceller employs the non-stationarity characteristics of the speech signal to estimate the relative transfer function and thus is difficult to apply when the noise is also non-stationary. Under the assumption of W-disjoint orthogonality between the speech and the non-stationary noise, the proposed algorithm finds the speech-dominant time-frequency bins of the input signal by inspecting the system output and the inter-microphone time delay. Only these bins are used to estimate the relative transfer function, so reliable estimates can be obtained under non-stationary noise conditions. The experimental results show that the proposed algorithm significantly improves the performance of the transfer-function generalized sidelobe canceller, while only sustaining a modest estimation error in adverse non-stationary noise environments.
Tang YINGJUN Xu DE Yang XU Liu QIFANG
We present a novel model named Integrated Latent Topic Model (ILTM), to learn and recognize natural scene category. Unlike previous work, which considered the discrepancy and common property separately among all categories, Our approach combines universal topics from all categories with specific topics from each category. As a result, the model is implemented to produce a few but specific topics and more generic topics among categories, and each category is represented in a different topics simplex, which correlates well with human scene understanding. We investigate the classification performance with variable scene category tasks. The experiments have shown our model outperforms latent-space methods with less training data.
Naohiko IWAKIRI Takehiko KOBAYASHI
This paper proposes an ultra-wideband double-directional spatio-temporal channel sounding technique using transformation between frequency- and time-domain (FD and TD) signals. Virtual antenna arrays, composed of omnidirectional antennas and scanners, are used for transmission and reception in the FD. After Fourier transforming the received FD signals to TD ones, time of arrival (TOA) is estimated using a peak search over the TD signals, and then angle of arrivals (AOA) and angle of departure (AOD) are estimated using a weighted angle histogram with a multiple signal classification (MUSIC) algorithm applied to the FD signals, inverse-Fourier transformed from the TD signals divided into subregions. Indoor channel sounding results validated that an appropriate weighting reduced a spurious level in the angle histogram by a factor of 0.1 to 0.2 in comparison with that of non-weighting. The proposed technique successfully resolved dominant multipath components, including a direct path, a single reflection, and a single diffraction, in line-of-sight (LOS) and non-LOS environments. Joint TOA and AOA/AOD spectra were also derived from the sounding signals. The spectra illustrated the dominant multipath components (agreed with the prediction by ray tracing) as clusters.
Eunju HWANG Yong Hyun LEE Kyung Jae KIM Jung Je SON Bong Dae CHOI
The IEEE 802.16e standard specifies the sleep mode and the idle mode of a mobile station (MS) for power saving. In this paper, to reduce the energy consumption of the MS, we employ the sleep mode while the MS is on-session, and the idle mode while it is off-session. Under the assumption that the time duration from the end of a session to the arrival of a new downlink session request follows an exponential distribution of the mean and that arrivals of messages during an on-session follow a Poisson process with rate λ, we analyze the awake mode period and the sleep mode period by using the busy period analysis of the M/G/1 queue, and then we derive the total mean length of an on-session which consists of a geometric number of awake mode periods and sleep mode periods. Since the sum of an on-session and an off-session constitutes a cycle, we can express the average power consumption in terms of the mean lengths of an awake mode period, a sleep mode period and an idle mode period. The average power consumption indicates how much the MS can save energy by employing the sleep mode and the idle mode. We also derive the Laplace Stieltjes transform (and the mean) of the queueing delay of messages to examine a tradeoff between the power consumption and the delay of messages. Analytical results, which are shown to be well-matched by simulations, address that our employment of the sleep mode and the idle mode provides a considerable reduction in the energy consumption of the MS.
Tacksung CHOI Sunkuk MOON Young-cheol PARK Dae-hee YOUN Seokpil LEE
In this paper, we propose a new feature selection algorithm for multi-class classification. The proposed algorithm is based on Gaussian mixture models (GMMs) of the features, and it uses the distance between the two least separable classes as a metric for feature selection. The proposed system was tested with a support vector machine (SVM) for multi-class classification of music. Results show that the proposed feature selection scheme is superior to conventional schemes.