the summation CFAR detector is widely used for the detection of narrowband signals. the normalized detection threshold determines the detection performance and must be appropriately chosen in practical applications. H...
详细信息
ISBN:
(纸本)9780889866706
the summation CFAR detector is widely used for the detection of narrowband signals. the normalized detection threshold determines the detection performance and must be appropriately chosen in practical applications. However, numerical problems often occur in the theoretical computation of the normalized detection threshold, particularly when channel power estimates are summed over a large number of overlapped input data blocks. this paper shows that the correlation between power estimates at the output of an FFT filter bank for successive input data blocks can be neglected for most common windows when the overlap ratio is less than or equal to 1/2. Under this constraint the normalized detection thresholds computed for overlapped and non-overlapped input data blocks are practically identical and the results for the probability of false alarm for a given threshold derived for non-overlapped input data blocks are applicable to overlapped input data blocks. this substantially simplifies the problem of computing the normalized detection threshold for overlapped input data blocks.
this paper describes a mathematically closed form transfer function of low delay maximally flat bandpass finite impulse response (FIR) filters. It is possible to obtain the transfer function withthe specified degree ...
详细信息
ISBN:
(纸本)9780889866706
this paper describes a mathematically closed form transfer function of low delay maximally flat bandpass finite impulse response (FIR) filters. It is possible to obtain the transfer function withthe specified degree of flatness at center frequency (ω = ±ω0) in the passband, ω=0 and ω=π. In addition, the obtained transfer function has the specified group delay at center frequency in the passband. the proposed transfer function has already contained the transfer function of the known linear phase maximally flat bandpass FIR filter. We show an effectiveness of the proposed transfer function by examples.
Data mining applications have facilitated knowledge discovery in a multitude of areas including marketing, biological and geological sciences, and image processing. through well established mathematical algorithms rel...
详细信息
ISBN:
(纸本)9780889867055
Data mining applications have facilitated knowledge discovery in a multitude of areas including marketing, biological and geological sciences, and image processing. through well established mathematical algorithms related to patternrecognition, predictive modeling, and anomaly detection, warehouses of data have been explored and interpreted withthe purposes of helping researchers unveil and understand complex patterns. In this paper, we present the development of a web-based Data Mining Utility, created withthe ultimate goal of allowing medical researchers to detect new artifacts that have been previously obscured through conventional signalprocessing techniques in neurological datasets. this web tool is powered by Data to Knowledge, which was created by the Automated Learning Group of the National Center for Supercomputing applications at the University of Illinois, by integrating a vast library of customizable data mining techniques into a unified programming environment. Using this framework, along withthe D2K Web Service, this Data Mining Utility will enhance an established Multi-Site Pediatric Network for fMRI in Childhood Epilepsy, as a means for researchers to gain further insight on medical case studies.
the Hilbert Transform (HT) and the analytic signal (AS) are widely used in their one-dimensional version for various applications. However, in the bi-dimensional (2D) case as occur for images, the definition of the 2D...
详细信息
ISBN:
(纸本)9783540742586
the Hilbert Transform (HT) and the analytic signal (AS) are widely used in their one-dimensional version for various applications. However, in the bi-dimensional (2D) case as occur for images, the definition of the 2D-HT is not unique and several approaches to it have been developed, having as one of the main goals to obtain a meaningful 2D-AS or analytic image, which can be used for various practical applications. In this work, one particular approach to the 2D-HT is introduced that allowed the calculation of analytic images which satisfy the basic properties that these functions have in the ID case, and that produces a 2D spectrum equal to zero in one quadrant. the methods for calculation of the discrete version of the 2D-HT and the associated AS are presented and analyzed, as well as two applications, for edge detection and for envelope detection in a 2D AM modulated radial chirp.
this paper describes initial results from a novel low dimensional surface parameterisation approach based on a modified Iterative Closest Point (ICP) registration process which uses vertex based Principal Component An...
详细信息
Linear subspace method based on principal component analysis has been applied for patternrecognition of high-dimensional data. However, the linear subspace method can not represent nonlinear characteristics of the da...
详细信息
ISBN:
(纸本)9780889865839
Linear subspace method based on principal component analysis has been applied for patternrecognition of high-dimensional data. However, the linear subspace method can not represent nonlinear characteristics of the data-distribution efficiently. In order to overcome this problem, some nonlinear subspace methods have been proposed. In this paper, we propose a novel nonlinear subspace method for patternrecognition using multi-layered perceptrons which can hierarchically construct a nonlinear subspace from the data-distribution. We introduce the optimization scheme of the subspace taking into account the classification. the proposed method achieves high classification accuracy by the optimization scheme and provides the finite subspace to avoid the crossover of the subspaces. We examine its effectiveness through some experiments.
this paper introduces SAFIA: sound source Segregation based on estimating incident Angle of each Frequency component of Input sources Acquired by multiple microphones, which utilizes the level differences of direction...
详细信息
ISBN:
(纸本)0889865477
this paper introduces SAFIA: sound source Segregation based on estimating incident Angle of each Frequency component of Input sources Acquired by multiple microphones, which utilizes the level differences of directional microphones located at the same place, namely aggregated microphones. SAFIA can realize sound source separation based on the difference of the power level between two microphones. In the conventional SAFIA, one microphone for target signal (handset) is located near the target signal. the other microphone for noise signal is located far from the target signal. Hence, SAFIA needs the large system size. Our research target is to solve this problem utilizing aggregated microphones. In this paper, we firstly explain the conventional SAFIA system and the problem of SAFIA. We then apply SAFIA to the aggregated microphones and compare the proposed method withthe conventional SAFIA. We demonstrate the experimental results of sound separation utilizing the proposed method.
A double-digit text-dependent speaker verification and text validation system is presented for use in telephone services. the system utilizes concatenated phoneme HMMs for both speech recognition and user authenticati...
详细信息
ISBN:
(纸本)0889865477
A double-digit text-dependent speaker verification and text validation system is presented for use in telephone services. the system utilizes concatenated phoneme HMMs for both speech recognition and user authentication, and works in a soundprompted mode. Tests with Hidden Markov Models (HMMs) using Perceptual Linear Prediction (PLP) and Mel Frequency Cepstral Coefficients (MFCC) as well as Cepstral Mean Subtraction (CMS) are performed to assess their effect on recognition performance. the paper also studies the effects of various factors such as the length of the training data, the number of embedded re-estimations and Gaussian mixtures in training of the HMMs, the use of world models, bootstrapping, and user-depended thresholds on the performance of speech recognition and speaker verification.
the proceedings contain 69 papers. the topics discussed include: notes on the passive and active redundancies in digital technology;design issues and implementation strategies for building on-chip voltage level-shifti...
详细信息
ISBN:
(纸本)9780889866072
the proceedings contain 69 papers. the topics discussed include: notes on the passive and active redundancies in digital technology;design issues and implementation strategies for building on-chip voltage level-shifting circuits;a novel image compression algorithm for hardware implementation;the study on the parameters used in calculating correlative dimension;an algorithm for tracking motion objects;multi-resolution wavelet analyses for detection of potential defects in motor bearings using vibration signals;factorization approach to control design - utilization for systems with parametric uncertainty;realization of the third-order high-pass transfer function;an evolutionary time-series model for partitioning a circuit pertaining to VLSI design using neuro-memetic algorithm;a novel duration model for speech recognition;Markov family models and its applications;and a processing element for digital asynchronous/synchronous vision chip.
the proceedings contain 90 papers. the topics discussed include: applying multiple instances and matching algorithm in face verification system;autonomous cropping and feature extraction using time-frequency marginal ...
ISBN:
(纸本)0889865833
the proceedings contain 90 papers. the topics discussed include: applying multiple instances and matching algorithm in face verification system;autonomous cropping and feature extraction using time-frequency marginal distributions for LPI radar classification;object recognition using don't care dimensions and its application to information retrieval system on cellular phone;comparison between weighted D-KNN and other classifiers for mandarin speech emotion recognition;quasi-motion extraction utilizing edge field analysis;an improved algorithm of viewpoint space partition in 3D object recognition application;experiments with soft fusion methods when bagging k-NN Classifiers;a nonlinear subspace method for patternrecognition using a nonlinear PCA;detecting the degree of anomaly in security videos by using a spatio-temporal feature of change;and illuminant-invariant Bayesian detection of moving video objects.
暂无评论