The theory of set membership (SM) identification is formulated, and applied to linear prediction (LP) analysis of speech. The LP parameters of a simulated vowel are identified as an illustration. The SM strategy resul...
详细信息
The theory of set membership (SM) identification is formulated, and applied to linear prediction (LP) analysis of speech. The LP parameters of a simulated vowel are identified as an illustration. The SM strategy results in a significant computational savings due to rejection of data which are informationless in the SM sense.
A method for recognition of speech of the nonverbal is presented. Results of the application of the procedure to simple utterances by cerebral palsied individuals and implications for an overall communications device ...
详细信息
A method for recognition of speech of the nonverbal is presented. Results of the application of the procedure to simple utterances by cerebral palsied individuals and implications for an overall communications device are addressed.
Fundamental issues related to the design of an intelligent communication (speech or text) device for the nonverbal, profoundly motor disabled are discussed. The device is to assist a person in the generation of a mess...
ISBN:
(纸本)9780818608117
Fundamental issues related to the design of an intelligent communication (speech or text) device for the nonverbal, profoundly motor disabled are discussed. The device is to assist a person in the generation of a message by suggesting message elements one-at-a-time, for yes/no response. A graph theoretic approach to represent contextual knowledge about a fixed language is proposed, and message generation is viewed as one of decentralized control of graph search. A strategy for integrating information from a speech recognition system is formalized. Its significance is to allow the user to attempt to speak the message. The search strategy is based on a parsing tree traversal technique which dynamically adjusts its message element presentations in accordance with contextual knowledge embedded in the graph, the information from speech recognition, and the user's responses to interrogation.
The concept of fast KL transform coding introduced earlier [7], [8] for first-order Markov processes and certain random fields has been extended to higher order autoregressive (AR) sequences and practical images yield...
详细信息
The concept of fast KL transform coding introduced earlier [7], [8] for first-order Markov processes and certain random fields has been extended to higher order autoregressive (AR) sequences and practical images yielding what we call recursive block coding (RBC) algorithms. In general, the rate-distortion performance for these algorithms is significantly superior to that of the conventional block KL transform algorithm. Moreover, these algorithms permit the use of small size transforms, thereby removing the need for fast transforms and making the hardware implementation of such coders more appealing. This improved performance has been verified for practical image data and results in suppression of the block-boundary effect commonly observed in traditional transform coding techniques. This is illustrated by comparing RBC with cosine transform coding using both one- and twodimensional algorithms. Examples of RBC encoded images at various rates are given.
Although the magnitude of the discrete Fourier transform of a maximal-length shift-register sequence is flat, except for its value at zero frequency, the higher resolution spectral content given by the Fourier-series ...
详细信息
Although the magnitude of the discrete Fourier transform of a maximal-length shift-register sequence is flat, except for its value at zero frequency, the higher resolution spectral content given by the Fourier-series transform is highly erratic. This little-known fact is described, and its ramifications on fast Fourier transforms of one-digit-extended pseudo noise and zero-padded pseudo noise are explained.
The popular class of synchronizers that consist of a quadratic nonlinearity followed by a phase-lock loop is investigated, and it is shown that the optimum design of the quadratic transformation is characterized in te...
详细信息
The popular class of synchronizers that consist of a quadratic nonlinearity followed by a phase-lock loop is investigated, and it is shown that the optimum design of the quadratic transformation is characterized in terms of a spectral correlation function for the signal to be synchronized to. It is also shown that the SNR performance of this quadratic transformation, and the mean-square phase jitter of the phaselock loop are both characterized in terms of spectral correlation functions. The conditions under which the optimum quadratic transformations, for symbol synchronization of BPSK, QPSK, SQPSK, and MSK, and for carrier synchronization of BPSK, reduce to the well-known matched-filter-squarer are identified. In addition, the well-known zeromean-square-phase-jitter condition is generalized from PAM to all synchronizable signals, and is characterized in terms of the spectral correlation function. The low-SNR maximum-likelihood synchronizer for all quadratically synchronizable signals is characterized in terms of a multiplicity of maximum-SNR quadratic spectral-line generators. A closed form implementation in terms of a matched filter, squarer, and symbol-rate-synchronized averager is obtained for BPSK and QPSK signals.
This paper discusses the relationship between the zero crossings or zeros of band-limited signals and their nonlinear transformations. It is proved that the bandwidth of a signal can be compressed by a ratio of 1/ n i...
详细信息
This paper discusses the relationship between the zero crossings or zeros of band-limited signals and their nonlinear transformations. It is proved that the bandwidth of a signal can be compressed by a ratio of 1/ n if and only if the signal has n th-order zero crossings or zeros (if complex). Also, a monotonic nonlinearity in the observation of a band-limited signal can be identified from the zero crossings (or zeros) of the derivative of the observed signal. (The results are for one-dimensional signals. Extensions to two-dimensional signals remain to be addressed.)
A new algorithm for ray tracing parametric surface patches is presented. The method uses quasi-Newton iteration to solve for the ray/surface intersection and utilizes ray-to-ray coherence by using numerical informatio...
详细信息
Factor inverse matched filtering (FIMF) and factor inverse filtering (FIF) are signal processing techniques used to obtain desired signal responses. Both are especially useful procedures for ‘‘pulse‐compression’’...
详细信息
Factor inverse matched filtering (FIMF) and factor inverse filtering (FIF) are signal processing techniques used to obtain desired signal responses. Both are especially useful procedures for ‘‘pulse‐compression’’ processing and channel measurements. The theory is developed for a simple channel and known noise power spectral density so that comparison may be made with matched filtering. Expressions for the pulse‐compression energy gain, nonflatness loss NFL, and total performance are derived. The NFL is useful in selecting the best among practical pulse‐compression modulations, and with FIMF and FIF, has been used extensively since 1974 by the authors and their co‐workers in underwater acoustic propagation measurements and ocean acoustictomography.
A layer stripping procedure for solving three‐dimensional Schr?dinger equationinverse scattering problems is developed. This procedure operates by recursively reconstructing the Radon transform of the potential from ...
详细信息
A layer stripping procedure for solving three‐dimensional Schr?dinger equationinverse scattering problems is developed. This procedure operates by recursively reconstructing the Radon transform of the potential from the jump in the Radon transform of the scattered field at the wave front. This reconstructed potential is then used to propagate the wave front and scattered field differentially further into the support of the potential. The connections between this differential procedure and integral equation procedures are then illustrated by the derivations of two well known exact integral equation procedures using the Radon transform and a generalized Radon transform. These procedures, as well as the layer stripping procedure, are then reduced to the familiar Born approximation result for this problem by neglecting multiple scattering events. This illustrates the central role of the Radon transform in both exact and approximate inversion procedures.
暂无评论