By Topic

Neural Networks for Signal Processing [1991]., Proceedings of the 1991 IEEE Workshop

Date Sept. 30 1991-Oct. 1 1991

Filter Results

Displaying Results 1 - 25 of 64
  • Neural Networks for Signal Processing. Proceedings of the 1991 IEEE Workshop (Cat. No.91TH0385-5)

    Publication Year: 1991
    Request permission for commercial reuse | PDF file iconPDF (27 KB)
    Freely Available from IEEE
  • Restricted learning algorithm and its application to neural network training

    Publication Year: 1991, Page(s):131 - 140
    Cited by:  Patents (2)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (304 KB)

    The authors propose a new (semi)-optimization algorithm, called the restricted learning algorithm, for a nonnegative evaluating function which is 2 times continuously differentiable on a compact set Ω in RN. The restricted learning algorithm utilizes the maximal excluding regions which are newly derived, and is shown to converge to the global ∈-optimum in Ω. A ... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Experiments with temporal resolution for continuous speech recognition with multi-layer perceptrons

    Publication Year: 1991, Page(s):405 - 410
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (316 KB)

    Previous work by the authors focused on the integration of multilayer perceptrons (MLP) into hidden Markov models (HMM) and on the use of perceptual linear prediction (PLP) parameters for the feature inputs to such nets. The system uses the Viterbi algorithm for temporal alignment. This algorithm is a simple and optimal procedure, but it necessitates a frame-based analysis in which all features ha... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A parallel learning filter system that learns the KL-expansion from examples

    Publication Year: 1991, Page(s):121 - 130
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (384 KB)

    A new method for learning in a single-layer linear neural network is investigated. It is based on an optimality criterion that maximizes the information in the outputs and simultaneously concentrates the outputs. The system consists of a number of so-called basic units and it is shown that the stable states of these basic units correspond to the (pure) eigenvectors of the input correlation matrix.... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A hybrid continuous speech recognition system using segmental neural nets with hidden Markov models

    Publication Year: 1991, Page(s):347 - 356
    Cited by:  Papers (2)  |  Patents (1)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (448 KB)

    The authors present the concept of a `segmental neural net' (SNN) for phonetic modeling in continuous speech recognition (CSR) and demonstrate how than can be used with a multiple hypothesis (or N-Best) paradigm to combine different CSR systems. In particular, they have developed a system that combines the SNN with a hidden Markov model (HMM) system. They believe that this is the first system inco... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Word recognition based on the combination of a sequential neural network and the GPDM discriminative training algorithm

    Publication Year: 1991, Page(s):376 - 384
    Cited by:  Papers (1)  |  Patents (1)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (280 KB)

    The authors propose an isolated-word recognition method based on the combination of a sequential neural network and a discriminative training algorithm using the Generalized Probabilistic Descent Method (GPDM). The sequential neural network deals with the temporal variation of speech by dynamic programming, and the GPDM discriminative training algorithm is used to discriminate easily confused word... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Fuzzy tracking of multiple objects

    Publication Year: 1991, Page(s):589 - 592
    Cited by:  Papers (1)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (160 KB)

    The authors have applied a previously developed MLANS neural network to the problem of tracking multiple objects in heavy clutter. In their approach the MLANS performs a fuzzy classification of all objects in multiple frames in multiple classes of tracks and random clutter. This novel approach to tracking using an optimal classification algorithm results in a dramatic improvement of performance: t... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Discriminative multi-layer feed-forward networks

    Publication Year: 1991, Page(s):11 - 20
    Cited by:  Papers (20)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (408 KB)

    The authors propose a new family of multi-layer, feed-forward network (FFN) architectures. This framework allows examination of several feed-forward networks, including the well-known multi-layer perceptron (MLP) network, the likelihood network (LNET) and the distance network (DNET), in a unified manner. They then introduce a novel formulation which embeds network parameters into a functional form... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Speech recognition using time-warping neural networks

    Publication Year: 1991, Page(s):337 - 346
    Cited by:  Patents (45)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (396 KB)

    The author proposes a time-warping neural network (TWNN) for phoneme-based speech recognition. The TWNN is designed to accept phonemes with arbitrary duration, whereas conventional phoneme recognition networks have a fixed-length input window. The purpose of this network is to cope with not only variability of phoneme duration but also time warping in a phoneme. The proposed network is composed of... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • On adaptive acquisition of spoken language

    Publication Year: 1991, Page(s):422 - 431
    Cited by:  Patents (13)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (400 KB)

    At present, automatic speech recognition technology is based upon constructing models of the various levels of linguistic structure assumed to compose spoken language. These models are either constructed manually or automatically trained by example. A major impediment is the cost, or even the feasibility, of producing models of sufficient fidelity to enable the desired level of performance. The pr... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Three-dimensional structured networks for matrix equation solving

    Publication Year: 1991, Page(s):80 - 89
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (380 KB)

    Structured networks are feedforward neural networks with linear neurons than use special training algorithms. Two three-dimensional (3-D) structured networks are developed for solving linear equations and the Lyapunov equation. The basic idea of the structured network approaches is: first, represent a given equation-solving problem by a 3-D structured network so that if the network matches a desir... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A neural network pre-processor for multi-tone detection and estimation

    Publication Year: 1991, Page(s):580 - 588
    Cited by:  Papers (1)  |  Patents (1)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (296 KB)

    A parallel bank of neural networks each trained in a specific band of the spectrum is proposed as a pre-processor for the detection and estimation of multiple sinusoids at low SNRs. A feedforward neural network model in the autoassociative mode, trained using the backpropagation algorithm, is used to construct this sectionized spectrum analyzer. The key concept behind this scheme is that, the netw... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Learned representation normalization: attention focusing with multiple input modules

    Publication Year: 1991, Page(s):111 - 120
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (524 KB)

    A large, multi-modular neural network can be envisaged for use in a complex, multi-task application. The optimum data representation for each sub-task of such an application is often unknown and different from the optimum data representation for the other sub-tasks. A method is needed that allows a network that contains several alternate input representations to learn to focus its attention on the... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A relaxation neural network model for optimal multi-level image representation by local-parallel computations

    Publication Year: 1991, Page(s):473 - 482
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (828 KB)

    A relaxation neural network model is proposed to solve the multi-level image representation problem by energy minimization in local and parallel computations. This network iteratively minimizes the computational energy defined by the local error in neighboring picture elements. This optimization method can generate high quality binary and multi-level images depending on local features, and can be ... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Nonlinear resampling transformation for automatic speech recognition

    Publication Year: 1991, Page(s):319 - 326
    Cited by:  Papers (3)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (260 KB)

    A new technique for speech signal processing called nonlinear resampling transformation (NRT) is proposed. The representation of a speech pattern derived from this technique has two important features: first, it reduces redundancy; second, it effectively removes the nonlinear variations of speech signals in time. The authors have applied NRT to the TI isolated-word database achieving a 99.66% reco... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Neural-network architecture for linear and nonlinear predictive hidden Markov models: application to speech recognition

    Publication Year: 1991, Page(s):411 - 421
    Cited by:  Papers (4)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (476 KB)

    A speech recognizer is developed using a layered neural network to implement speech-frame prediction and using a Markov chain to modulate the network's weight parameters. The authors postulate that speech recognition accuracy is closely linked to the capability of the predictive model in representing long-term temporal correlations in data. Analytical expressions are obtained for the correlation f... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Probability estimation by feed-forward networks in continuous speech recognition

    Publication Year: 1991, Page(s):309 - 318
    Cited by:  Papers (6)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (436 KB)

    The authors review the use of feedforward neural networks as estimators of probability densities in hidden Markov modelling. In this paper, they are mostly concerned with radial basis functions (RBF) networks. They not the isomorphism of RBF networks to tied mixture density estimators; additionally they note that RBF networks are trained to estimate posteriors rather than the likelihoods estimated... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • An alternative proof of convergence for Kung-Diamantaras APEX algorithm

    Publication Year: 1991, Page(s):40 - 49
    Cited by:  Papers (2)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (276 KB)

    The problem of adaptive principal components extraction (APEX) has gained much interest. In 1990, a new neuro-computation algorithm for this purpose was proposed by S. Y. Kung and K. I. Diamautaras. (see ICASSP 90, p.861-4, vol.2, 1990). An alternative proof is presented to illustrate that the K-D algorithm is in fact richer than has been proved before. The proof shows that the neural network will... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Vector quantization of images using neural networks and simulated annealing

    Publication Year: 1991, Page(s):552 - 561
    Cited by:  Papers (1)  |  Patents (1)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (400 KB)

    Vector quantization (VQ) has already been established as a very powerful data compression technique. Specification of the `codebook', which contains the best possible collection of `codewords', effectively representing the variety of source vectors to be encoded is one of the most critical requirements of VQ systems, and belongs, for most applications, to the class of hard optimization problems. A... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Connectionist speaker normalization and its applications to speech recognition

    Publication Year: 1991, Page(s):357 - 366
    Cited by:  Papers (5)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (448 KB)

    Speaker normalization may have a significant impact on both speaker-adaptive and speaker-independent speech recognition. In this paper, a codeword-dependent neural network (CDNN) is presented for speaker normalization. The network is used as a nonlinear mapping function to transform speech data between two speakers. The mapping function is characterized by two important properties. First, the asse... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Word recognition with the feature finding neural network (FFNN)

    Publication Year: 1991, Page(s):289 - 298
    Cited by:  Papers (3)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (364 KB)

    An overview of the architecture and capabilities of the work recognizer FFNN (`feature finding neural network') is given. FFNN finds features in a self-organizing way which are relatively invariant in the presence of time distortions and changes in speaker characteristics. Fast and optimal feature selection rules have been developed to perform this task. With FFNN, essential problems of word recog... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A mapping approach for designing neural sub-nets

    Publication Year: 1991, Page(s):70 - 79
    Cited by:  Papers (2)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (300 KB)

    Several investigators have constructed back-propagation (BP) neural networks by assembling smaller, pre-trained building blocks. This approach leads to faster training and provides a known topology for the network. The authors carry this process down one additional level, by describing methods for mapping given functions to sub-blocks. First, polynomial approximations to the desired function are f... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Segment-based speaker adaptation by neural network

    Publication Year: 1991, Page(s):442 - 451
    Cited by:  Papers (3)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (376 KB)

    The authors propose a segment-to-segment speaker adaptation technique using a feed-forward neural network with a time shifted sub-connection architecture. Differences in voice individuality exist in both the spectral and temporal domains. It is generally known that frame based speaker adaptation techniques can not compensate for speaker individuality in the temporal domain. Segment based speaker a... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A time-derivative neural net architecture-an alternative to the time-delay neural net architecture

    Publication Year: 1991, Page(s):367 - 375
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (384 KB)

    Though the time-delay neural net architecture has been recently used in a number of speech recognition applications, it has the problem that it can not use longer temporal contexts because this increases the number of connection weights in the network. This is a serious bottleneck because the use of larger temporal contexts can improve the recognition performance. In this paper, a time-derivative ... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Configuring stack filters by the LMS algorithm

    Publication Year: 1991, Page(s):570 - 579
    Cited by:  Papers (6)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (368 KB)

    Stack filters are a class of sliding-window nonlinear digital filters that possess the weak superposition property (threshold decomposition) and the ordering property known as the stacking property. They have been demonstrated to be robust in suppressing noise. A new method based on the least means squares (LMS) algorithm is developed to adaptively configure a stack filter. Experimental results ar... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.