Predictive Connectionist Approach to Speech Recognition

Petek, B.

doi:10.1007/11520153_10

Cited by 2 publications

(2 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To this aim, several non linear speech processing techniques have been proposed, such as non linear parametric and non parametric autoregressive models, speech fluid dynamics, modulation, and fractal methods. Some of these methods are theoretically discussed in this volume by Bastari et al [3], Chollet et al [7], Faundez-Zanuy [24], Haykin [33], Kubin et al [43], Murphy and Akande [46], Petek [57], and Stylianou [81]. The results are very encouraging, showing that nonlinear speech processing methods offers a good alternative to conventional speech encoding techniques.…”

Section: Nonlinearities In Speech Acousticsmentioning

confidence: 79%

Some Notes on Nonlinearities of Speech

Esposito

Marinaro

2005

Nonlinear Speech Modeling and Applications

View full text Add to dashboard Cite

Abstract. Speech is exceedingly nonlinear. Efforts to propose non-linear models of its dynamics are worth to be made but difficult to implement since nonlinearity is not easily handled from an engineering and mathematical point of view. This paper is an attempt to make accessible to untrained people the notion of nonlinearity in speech, revising several nonlinear speech phenomena and the engineering endeavour for modeling them.

show abstract

Section: Nonlinearities In Speech Acousticsmentioning

confidence: 79%

Some Notes on Nonlinearities of Speech

Esposito

Marinaro

2005

Nonlinear Speech Modeling and Applications

View full text Add to dashboard Cite

show abstract

“…Its classification decision-making ability and descriptive power on uncertain information have already been accepted by the whole world, but its descriptive power on dynamic time signal is not entirely satisfactory. Generally speaking, ANN classifier can only solve statistic model classification problem, but it doesn t involve in sequence processing [11,12]. In order to overcome respective disadvantages of HMM and ANN, in recent years, many scholars attempt to combine HMM, which has stronger time correction function with ANN, which has stronger classification ability to further improve accuracy rate of speech recognition.…”

Section: Training Learning Algorithm After Improvement----hmm/pnn Hybmentioning

confidence: 99%

The study of speech training and learning method based on DIVA model

Zhang

Chenhong

2015

2015 34th Chinese Control Conference (CCC)

View full text Add to dashboard Cite

DIVA (Directions Into of Articulators) model is a kind of self-adaptive neutral network model which controls movements of a simulated vocal tract in order to produce words, syllables or phonemes. However, there exist poor classification ability, out of consideration of overlap and other deficiencies among multiple modeling primitives in current Hidden Markov(HMM) training algorithm. It impacts speech recognition rate of the model. Therefore, this paper proposes a hybrid model HMM/PNN, which is to use Predictive Neural Network (PNN) in ANN(Artificial neutral network) to calculate station posterior distribution of Hidden Markov Model. The acoustic model of DIVA is reconstructed through extracting acoustic parameter, choosing modeling unit and other methods. The simulations show that after training and learning the pronunciation of compound vowel by using new HMM/PNN model, there s not huge difference between the waveform of the acquired speech and that of real person, in addition, the recognition rate is also improved. All these verify the effectiveness and accuracy of this method.

show abstract

Predictive Connectionist Approach to Speech Recognition

Cited by 2 publications

References 20 publications

Some Notes on Nonlinearities of Speech

Some Notes on Nonlinearities of Speech

The study of speech training and learning method based on DIVA model

Contact Info

Product

Resources

About