Pub Date : 1997-11-02DOI: 10.1109/ACSSC.1997.680208
W. Barrett
Automated face recognition (AFR) has received increased attention. We describe two general approaches to the problem and discuss their effectiveness and robustness with respect to several possible applications. We also discuss some issues of run-time performance. The AFR technology falls into three main subgroups, which represent more-or-less independent approaches to the problem: neural network solutions, eigenface solutions, and wavelet/elastic matching solutions. Each of these first requires that a facial image be identified in a scene, a process called segmentation. The image should be normalized to some extent. Normalization is usually a combination of linear translation, rotation and scaling, although the elastic matching method includes spatial transformations.
{"title":"A survey of face recognition algorithms and testing results","authors":"W. Barrett","doi":"10.1109/ACSSC.1997.680208","DOIUrl":"https://doi.org/10.1109/ACSSC.1997.680208","url":null,"abstract":"Automated face recognition (AFR) has received increased attention. We describe two general approaches to the problem and discuss their effectiveness and robustness with respect to several possible applications. We also discuss some issues of run-time performance. The AFR technology falls into three main subgroups, which represent more-or-less independent approaches to the problem: neural network solutions, eigenface solutions, and wavelet/elastic matching solutions. Each of these first requires that a facial image be identified in a scene, a process called segmentation. The image should be normalized to some extent. Normalization is usually a combination of linear translation, rotation and scaling, although the elastic matching method includes spatial transformations.","PeriodicalId":240431,"journal":{"name":"Conference Record of the Thirty-First Asilomar Conference on Signals, Systems and Computers (Cat. No.97CB36136)","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-11-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127629637","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1997-11-02DOI: 10.1109/ACSSC.1997.680176
O. Besson, P. Stoica
This paper addresses the estimation of the center frequency of complex exponential signals with time-varying amplitude. A method, which requires few assumptions regarding the signal's envelope is proposed. It is based on the polar decomposition of a certain covariance matrix. The polar decomposition, a generalization to matrices of the complex number representation z=re/sup i/spl theta// with r>0, is particularly suitable for the application considered. The notion of truncated polar decomposition is introduced. Simple schemes for estimating the signal's frequency are presented, based on these decompositions. The methods presented herein do not rely on any assumed structure for the time-varying amplitude, and they are shown to possess good performance in a large class of signals. The effectiveness and robustness of our method is demonstrated on real radar data.
{"title":"On frequency estimation of exponential signals with time-varying amplitude via polar decomposition","authors":"O. Besson, P. Stoica","doi":"10.1109/ACSSC.1997.680176","DOIUrl":"https://doi.org/10.1109/ACSSC.1997.680176","url":null,"abstract":"This paper addresses the estimation of the center frequency of complex exponential signals with time-varying amplitude. A method, which requires few assumptions regarding the signal's envelope is proposed. It is based on the polar decomposition of a certain covariance matrix. The polar decomposition, a generalization to matrices of the complex number representation z=re/sup i/spl theta// with r>0, is particularly suitable for the application considered. The notion of truncated polar decomposition is introduced. Simple schemes for estimating the signal's frequency are presented, based on these decompositions. The methods presented herein do not rely on any assumed structure for the time-varying amplitude, and they are shown to possess good performance in a large class of signals. The effectiveness and robustness of our method is demonstrated on real radar data.","PeriodicalId":240431,"journal":{"name":"Conference Record of the Thirty-First Asilomar Conference on Signals, Systems and Computers (Cat. No.97CB36136)","volume":"39 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-11-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128849685","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1997-11-02DOI: 10.1109/ACSSC.1997.679117
A. Srivastava, D. Fuhrmann
Estimation of dynamic subspaces is important in blind-channel identification for multiuser wireless communications and active computer vision. Mathematically, a subspace can either be parameterized non-uniquely by a linearly-independent basis, or uniquely, by a projection matrix. We present a stochastic gradient technique for optimization on projective representations of subspaces. This technique is intrinsic, i.e. it utilizes the geometry of underlying parameter space (Grassman manifold) and constructs gradient flows on the manifold for local optimization. The addition of a stochastic component to the search process guarantees global minima and a discrete jump component allows for uncertainty in rank of the subspace (simultaneous model order estimation).
{"title":"Gradient flows on projection matrices for subspace estimation","authors":"A. Srivastava, D. Fuhrmann","doi":"10.1109/ACSSC.1997.679117","DOIUrl":"https://doi.org/10.1109/ACSSC.1997.679117","url":null,"abstract":"Estimation of dynamic subspaces is important in blind-channel identification for multiuser wireless communications and active computer vision. Mathematically, a subspace can either be parameterized non-uniquely by a linearly-independent basis, or uniquely, by a projection matrix. We present a stochastic gradient technique for optimization on projective representations of subspaces. This technique is intrinsic, i.e. it utilizes the geometry of underlying parameter space (Grassman manifold) and constructs gradient flows on the manifold for local optimization. The addition of a stochastic component to the search process guarantees global minima and a discrete jump component allows for uncertainty in rank of the subspace (simultaneous model order estimation).","PeriodicalId":240431,"journal":{"name":"Conference Record of the Thirty-First Asilomar Conference on Signals, Systems and Computers (Cat. No.97CB36136)","volume":"51 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-11-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126425222","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1997-11-02DOI: 10.1109/ACSSC.1997.679193
W. Lu
This paper describes three wavelet-based methods for noise reduction of still images: (i) hyperbolic shrinkage with a level-dependent thresholding policy; (ii) hyperbolic shrinkage with a two-dimensional cross-validation-based thresholding; and (iii) block SVD-wavelet denoising. All three methods make use of hyperbolic shrinkage rather than conventional soft shrinkage. As the thresholding of wavelet coefficients is concerned, at each level of wavelet decomposition, the first method employs a level-dependent universal threshold determined by the coefficient variance and the number of the coefficients at that level; while the second method extends Nason's (1994) cross-validation approach to the 2-D case. In the third method, an image is divided into several subimages (blocks) and singular value decomposition (SVD) is applied to each block. The singular values obtained are then truncated and each pair of singular vectors are treated as 1-D noisy signals and are denoised using a wavelet-based method. The subimages are then reconstructed using the truncated singular values and denoised singular vectors.
{"title":"Wavelet approaches to still image denoising","authors":"W. Lu","doi":"10.1109/ACSSC.1997.679193","DOIUrl":"https://doi.org/10.1109/ACSSC.1997.679193","url":null,"abstract":"This paper describes three wavelet-based methods for noise reduction of still images: (i) hyperbolic shrinkage with a level-dependent thresholding policy; (ii) hyperbolic shrinkage with a two-dimensional cross-validation-based thresholding; and (iii) block SVD-wavelet denoising. All three methods make use of hyperbolic shrinkage rather than conventional soft shrinkage. As the thresholding of wavelet coefficients is concerned, at each level of wavelet decomposition, the first method employs a level-dependent universal threshold determined by the coefficient variance and the number of the coefficients at that level; while the second method extends Nason's (1994) cross-validation approach to the 2-D case. In the third method, an image is divided into several subimages (blocks) and singular value decomposition (SVD) is applied to each block. The singular values obtained are then truncated and each pair of singular vectors are treated as 1-D noisy signals and are denoised using a wavelet-based method. The subimages are then reconstructed using the truncated singular values and denoised singular vectors.","PeriodicalId":240431,"journal":{"name":"Conference Record of the Thirty-First Asilomar Conference on Signals, Systems and Computers (Cat. No.97CB36136)","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-11-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121181620","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1997-11-02DOI: 10.1109/ACSSC.1997.680562
K. Kongelbeck
This paper develops a digital method of in phase (I) and quadrature (Q) conversion using techniques similar to those in the classical analog method. It uses a graphical method of representing the spectra of the processed signals that gives an intuitive feel for the process. It shows that while a sample rate of twice the bandwidth of the input signal is required, the computational complexity is reduced by a higher sample rate. The graphical technique gives insight into the filtering, IF frequency and sample rate requirements.
{"title":"A spectral method of digital I Q conversion","authors":"K. Kongelbeck","doi":"10.1109/ACSSC.1997.680562","DOIUrl":"https://doi.org/10.1109/ACSSC.1997.680562","url":null,"abstract":"This paper develops a digital method of in phase (I) and quadrature (Q) conversion using techniques similar to those in the classical analog method. It uses a graphical method of representing the spectra of the processed signals that gives an intuitive feel for the process. It shows that while a sample rate of twice the bandwidth of the input signal is required, the computational complexity is reduced by a higher sample rate. The graphical technique gives insight into the filtering, IF frequency and sample rate requirements.","PeriodicalId":240431,"journal":{"name":"Conference Record of the Thirty-First Asilomar Conference on Signals, Systems and Computers (Cat. No.97CB36136)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-11-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121801115","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1997-11-02DOI: 10.1109/ACSSC.1997.679182
S. Kinjo, M. Oshiro, H. Ochi, M. Nayeri
A frequency-domain adaptive filter (FDAF) with a parallel structure and reduced computational complexity for precise adaptive system identification has been proposed. However, the presence of additive noise degrades the performance of the FDAF, and the degradation is conspicuous in the worse case input signal. We propose a new FDAF which improves the performance of the conventional FDAF for system identification in the presence of additive noise and colored input signal.
{"title":"A robust frequency-domain adaptive filter with colored input signal","authors":"S. Kinjo, M. Oshiro, H. Ochi, M. Nayeri","doi":"10.1109/ACSSC.1997.679182","DOIUrl":"https://doi.org/10.1109/ACSSC.1997.679182","url":null,"abstract":"A frequency-domain adaptive filter (FDAF) with a parallel structure and reduced computational complexity for precise adaptive system identification has been proposed. However, the presence of additive noise degrades the performance of the FDAF, and the degradation is conspicuous in the worse case input signal. We propose a new FDAF which improves the performance of the conventional FDAF for system identification in the presence of additive noise and colored input signal.","PeriodicalId":240431,"journal":{"name":"Conference Record of the Thirty-First Asilomar Conference on Signals, Systems and Computers (Cat. No.97CB36136)","volume":"35 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-11-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124921914","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1997-11-02DOI: 10.1109/ACSSC.1997.680183
S. Barbarossa, A. Scaglione
In this paper we propose a method for motion compensation and target classification, where the feature extraction is performed directly on the received signal. The method does not rely upon the formation of a radar image for classification, as opposed to conventional techniques. The proposed procedure assumes that the target is a rigid body and possesses at least four dominant scatterers, whose instantaneous frequencies can be tracked separately. No assumptions are made on the relative radar-target motion which can be absolutely arbitrary. To discriminate echoes superimposed in time and, possibly, in frequency, we map the received signals onto the time-frequency plane and then apply a parametric estimation method. From a set of at least four dominant echoes, we extract a set of features which are independent of target location and aspect angle. Some simulation results are shown to validate the proposed procedure.
{"title":"Motion compensation and target classification based on parametric modeling of the instantaneous frequency of echoes backscattered from rigid bodies","authors":"S. Barbarossa, A. Scaglione","doi":"10.1109/ACSSC.1997.680183","DOIUrl":"https://doi.org/10.1109/ACSSC.1997.680183","url":null,"abstract":"In this paper we propose a method for motion compensation and target classification, where the feature extraction is performed directly on the received signal. The method does not rely upon the formation of a radar image for classification, as opposed to conventional techniques. The proposed procedure assumes that the target is a rigid body and possesses at least four dominant scatterers, whose instantaneous frequencies can be tracked separately. No assumptions are made on the relative radar-target motion which can be absolutely arbitrary. To discriminate echoes superimposed in time and, possibly, in frequency, we map the received signals onto the time-frequency plane and then apply a parametric estimation method. From a set of at least four dominant echoes, we extract a set of features which are independent of target location and aspect angle. Some simulation results are shown to validate the proposed procedure.","PeriodicalId":240431,"journal":{"name":"Conference Record of the Thirty-First Asilomar Conference on Signals, Systems and Computers (Cat. No.97CB36136)","volume":"101 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-11-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122403629","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1997-11-02DOI: 10.1109/ACSSC.1997.679126
M. Ho
Previous advances using iterative (turbo) decoding techniques have resulted in codes that not only approach the Shannon limit for additive white Gaussian noise (AWGN) channels but are also readily implemented in practice. The original turbo codes include a pair of encoders separated by an interleaver and concatenated in parallel. New turbo codes developed by Benedetto et al. (see IEEE Trans. on. Communications, vol.44, no.5, p.591-600, 1996) include a pair of convolutional encoders also separated by an interleaver but concatenated serially. We extend the work of Benedetto et al. to serially-concatenated trellis-coded modulation (SCTCM). We determine the Bhattacharyya bound for maximum-likelihood decoding using the concept of the uniform interleaver. This bound enables us to determine the key parameters that can be used to develop a systematic code design strategy. We discuss both these results and the accuracy of the bound.
{"title":"Performance bounds for serially-concatenated trellis-coded modulation","authors":"M. Ho","doi":"10.1109/ACSSC.1997.679126","DOIUrl":"https://doi.org/10.1109/ACSSC.1997.679126","url":null,"abstract":"Previous advances using iterative (turbo) decoding techniques have resulted in codes that not only approach the Shannon limit for additive white Gaussian noise (AWGN) channels but are also readily implemented in practice. The original turbo codes include a pair of encoders separated by an interleaver and concatenated in parallel. New turbo codes developed by Benedetto et al. (see IEEE Trans. on. Communications, vol.44, no.5, p.591-600, 1996) include a pair of convolutional encoders also separated by an interleaver but concatenated serially. We extend the work of Benedetto et al. to serially-concatenated trellis-coded modulation (SCTCM). We determine the Bhattacharyya bound for maximum-likelihood decoding using the concept of the uniform interleaver. This bound enables us to determine the key parameters that can be used to develop a systematic code design strategy. We discuss both these results and the accuracy of the bound.","PeriodicalId":240431,"journal":{"name":"Conference Record of the Thirty-First Asilomar Conference on Signals, Systems and Computers (Cat. No.97CB36136)","volume":"45 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-11-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116499229","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1997-11-02DOI: 10.1109/ACSSC.1997.680178
F. Gini, G. Giannakis
Parameter estimation for a combination of a polynomial phase signal (PPS) and a hyperbolic frequency modulation (FM) is addressed. A novel approach is proposed that allows one to decouple estimation of the FM parameters from that of the PPS, exploiting the properties of the multi-lag high-order ambiguity function (ml-HAF). The accuracy achievable by any unbiased estimator of the hybrid FM-PPS parameters is investigated by means of the Cramer-Rao lower bounds (CRLB). Performance analysis is carried out and the CRLBs are compared with simulation results.
{"title":"Parameter estimation of hybrid hyperbolic FM and polynomial phase signals using the multi-lag high-order ambiguity function","authors":"F. Gini, G. Giannakis","doi":"10.1109/ACSSC.1997.680178","DOIUrl":"https://doi.org/10.1109/ACSSC.1997.680178","url":null,"abstract":"Parameter estimation for a combination of a polynomial phase signal (PPS) and a hyperbolic frequency modulation (FM) is addressed. A novel approach is proposed that allows one to decouple estimation of the FM parameters from that of the PPS, exploiting the properties of the multi-lag high-order ambiguity function (ml-HAF). The accuracy achievable by any unbiased estimator of the hybrid FM-PPS parameters is investigated by means of the Cramer-Rao lower bounds (CRLB). Performance analysis is carried out and the CRLBs are compared with simulation results.","PeriodicalId":240431,"journal":{"name":"Conference Record of the Thirty-First Asilomar Conference on Signals, Systems and Computers (Cat. No.97CB36136)","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-11-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127707716","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 1997-11-02DOI: 10.1109/ACSSC.1997.680582
A. Aziz, M. Tummala, R. Cristi
The problem of decision fusion in distributed sensor systems is considered. Distributed sensors pass their decisions to a fusion center that combines the received decisions from the various sensors into a final global decision. The case where only two sensors are combined using an AND fusion rule was analyzed by Kovattana (1973), and the cases where two and three sensors are combined using AND and OR fusion rules were analyzed by Fefjar (1978). Stearns (1983) considered the case of combining two sensors using AND and OR fusion rules and showed that the receiver operating characteristics of the AND and the OR combiners must be intersected, AND was superior to OR at low false-alarm probabilities, and OR was superior to AND at high false-alarm probabilities. This paper shows that the optimal fusion rule does not only depend on the desired false alarm probability and the signal to noise ratio but also depends on the probability distribution function.
{"title":"Optimal data fusion strategies using multiple-sensor detection systems","authors":"A. Aziz, M. Tummala, R. Cristi","doi":"10.1109/ACSSC.1997.680582","DOIUrl":"https://doi.org/10.1109/ACSSC.1997.680582","url":null,"abstract":"The problem of decision fusion in distributed sensor systems is considered. Distributed sensors pass their decisions to a fusion center that combines the received decisions from the various sensors into a final global decision. The case where only two sensors are combined using an AND fusion rule was analyzed by Kovattana (1973), and the cases where two and three sensors are combined using AND and OR fusion rules were analyzed by Fefjar (1978). Stearns (1983) considered the case of combining two sensors using AND and OR fusion rules and showed that the receiver operating characteristics of the AND and the OR combiners must be intersected, AND was superior to OR at low false-alarm probabilities, and OR was superior to AND at high false-alarm probabilities. This paper shows that the optimal fusion rule does not only depend on the desired false alarm probability and the signal to noise ratio but also depends on the probability distribution function.","PeriodicalId":240431,"journal":{"name":"Conference Record of the Thirty-First Asilomar Conference on Signals, Systems and Computers (Cat. No.97CB36136)","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"1997-11-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127825346","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：