Safdar M. Asghar

Matrix quantization with vector quantization error compensation and neural network postprocessing for robust speech recognition

A speech recognition system utilizes both matrix and vector quantizers as front ends to a second stage speech classifier such as hidden Markov models (HMMs) and utilizes neural network postprocessing to, for example, improve speech recognition performance. Matrix quantization exploits the “evolution” of the speech short-term spectral envelopes as well as frequency domain information, and vector quantization (VQ) primarily operates on frequency domain information. Time domain information may be substantially limited which may introduce error into the matrix quantization, and the VQ may provide error compensation. The matrix and vector quantizers may split spectral subbands to target selected frequencies for enhanced processing and may use fuzzy associations to develop fuzzy observation sequence data. A mixer provides a variety of input data to the neural network for classification determination. The neural networks ability to analyze the input data generally enhances recognition accuracy. Fuzzy operators may be utilized to reduce quantization error. Multiple codebooks may also be combined to form single respective codebooks for split matrix and split vector quantization to reduce processing resources demand.

Speech recognition system having a quantizer using a single robust codebook designed at multiple signal to noise ratios

In one embodiment, a speech recognition system is organized with a fuzzy matrix quantizer with a single codebook representing u codewords. The single codebook is designed with entries from u codebooks which are designed with respective words at multiple signal to noise ratio levels. Such entries are, in one embodiment, centroids of clustered training data. The training data is, in one embodiment, derived from line spectral frequency pairs representing respective speech input signals at various signal to noise ratios. The single codebook trained in this manner provides a codebook for a robust front end speech processor, such as the fuzzy matrix quantizer, for training a speech classifier such as a u hidden Markov models and a speech post classifier such as a neural network. In one embodiment, a fuzzy Viterbi algorithm is used with the hidden Markov models to describe the speech input signal probabilistically.

Apparatus and method for analyzing speech signals to determine parameters expressive of characteristics of the speech signals

An apparatus and method for locating a plurality of roots of a line spectrum pair expression on a unit circle. The method comprises the steps of: (1) receiving an initial value for locating a first site on the unit circle; (2) receiving a step value for defining an arc-distance on the unit circle; (3) generating intervals on the unit circle, each having a lower limit and an upper limit; the lower limit of the initial interval is the initial value and the upper limit of the initial interval is displaced on the unit circle from the initial value by the arc-distance; each succeeding interval has its lower limit coincident with the upper limit of the next preceding interval and has its upper limit displaced on the unit circle from its lower limit by the arc-distance; (4) evaluating the expression for at least the upper limit and the lower limit of each respective interval; (5) recognizing a root of the expression when the expression changes sign within an interval; (6) designating each such interval as a solution interval; and (7) generating the lower and upper limits of each solution interval to identify where the root is located. The apparatus comprises a waveform generator which receives the initial value and the step value and defines the arc-distance, and generates the intervals; a zero detector for recognizing roots of the expression when the expression changes sign.