Ken'iti Kido

A new FFT algorithm of radix 3,6, and 12

A new algorithm for implementation of radix 3, 6, and 12 FFT is introduced. An FFT using this algorithm is computed in an ordinary (1,j) complex plane and the number of additions can be significantly reduced; the number of multiplication is also reduced. High efficiency of the algorithm is derived from the fact that, if an input sequence is favorably reordered, rotating factors can be treated in pairs so that the rotating factors are conjugate to each other.

Recognition of consonant based on the perceptron model

This paper proposes a new method for the recognition of consonant based on the Perceptron model. The recognition model is composed of the sensory, feature extraction, response and lateral inhibition layers. The recognition scores of 90.4% to 98.4% are obtained for unvoiced affricates, unvoiced plosives, unvoiced and voiced fricatives.

Accurate autoregressive spectrum estimation at low signal-to-noise ratio using a phase matching technique

This paper describes a new method of accurately estimating the parameters of an autoregressive (AR) process contaminated by high-level white noise. Based on the phase matching technique, it minimizes the difference between the phase of the all-zero model and the phase of the maximum phase signal reconstructed from the power spectrum of the observed signal. The parameters of the AR model are obtained from the finite length sequence of the estimated all-zero model. The proposed method works only when the order of the AR model is known a priori at present. However, since the phase matching technique satisfies the conditions needed to apply the least mean-square method, the AR parameters are estimated accurately even at a low signal-to-noise ratio. With the iterative or noniterative methods as discussed in the recent literature, it is not possible to reconstruct the all-zero model from the power spectrum when there are dips and peaks having no correlation with the poles of original AR signal in the power spectrum. The method proposed in this paper allows one to accurately reconstruct the phase from the power spectrum in such cases. Finally, it is confirmed with computer simulations and experiments that the proposed method is useful for accurate estimation of the AR parameters.

A new approach to time dependent AR modeling of signals and its application to analysis of the fourth heart sound

The authors present a method for estimating spectrum transition between short-length signals of succeeding frames in low-SNR cases when the transition pattern is complex and/or there are large differences in the transition patterns among the individual sets of multiframe signals. The present approach uses a linear algorithm without any basic functions. Instead, the authors use the spectrum transition constraint, and the singular value decomposition. (SVD)-based technique is applied to obtain more accurate estimates. For the analysis of multiframe signals of the fourth heart sounds obtained during a stress test, significant differences in the transition patterns are clearly detected in the spectra between patients with myocardial infarction and normal persons The significant characteristics of these transition patterns may be applied to acoustic diagnosis of heart disease. >

Spoken word recognition system for unlimited speakers

This paper describes the newly improved spoken word recognition system. The accuracy of the word recognition is improved by implementing new features extraction method and new phoneme connecting rules in addition to the previously reported system. The new features for the phoneme recognition are the spectral local peaks and four parameters by using the least squares fit line of speech spectrum. By increasing the features for the phoneme recognition, and by using new phoneme connecting rules, the accuracies of phoneme recognition and segmentation are improved. In the last step of the system, the item of the dictionary having maximum similarity to the recognized phonemic sequence is chosen. Every item of the dictionary is written in phonemic symbols, and it is easy to change the objective words. The score of the word recognition was found to be 85.7% for 166 city names uttered by 15 male speakers.

A Japanese text dictation system based on phoneme recognition and a dependency grammar

A prototype of a Japanese text dictation system has been developed. It is composed of an acoustic processor, a Bunsetsu-unit spotting processor, and a syntactic processor with semantic constraints. The acoustic processor is constructed using the modified LVQ2 method. The modified LVQ2 method achieves a high phoneme recognition performance of 86.1%. The syntax driven continuous DP matching algorithm is used for spotting Bunsetsu-units. This method greatly reduces the amount of computation and the storage capacity necessary for spotting the Bunsetsu units. Analysis of the dependency structure among the Bunsetsu-unit candidates is effectively carried out using the syntactic and semantic information.<<ETX>>

A speaker independent word recognition system based on phoneme recognition for a large size (212 words) vocabulary

This paper describes the speaker-independent spoken word recognition system for a large size vocabulary. Speech is analyzed by the filter bank, from whose logarithmic spectrum the 11 features are extracted every 10 ms. Using the features the speech is first segmented and the primary phoneme recognition is carried out for every segment using the Bayes decision method. After correcting errors in segmentation and phoneme recognition, the secondary recognition of part of the consonants is carried out and the phonemic sequence is determined. The word dictionary item having maximum likelihood to the sequence is chosen as the recognition output. The 75.9% score for the phoneme recognition and the 92.4% score for the word recognition are obtained for the training samples in the 212 words uttered by 10 male and 10 female speakers. For the same words uttered by 30 male and 20 female speakers different from the above speakers, the 88.1% word recognition score is obtained.

Devanagari character recognition using structure analysis

A method of character recognition using prior knowledge of the script is proposed. Devanagari, a script widely used in India at present, and found in Buddhist texts of the past, is used for this purpose. This study is confined to recognizing a particular font used in a printed Buddhist text: Saddharmapundarika.<<ETX>>

Spoken word recognition system for unlimited adult male speakers

An online automatic spoken word recognition system has been developed for the researches on the automatic recognition of speech. In this system, the spoken word is first frequency analysed with a filter bank of single tuned low selectivity filters. Three major local peaks in the spectrum and the amplitude of the speech wave are extracted every 10 ms. The frequencies of two local peaks are used for classifying the vowels, and the frequencies of three local peaks, the movements of them and the amplitude are used for classifying the semi-vowels and consonants. Input speech is thus transformed into a sequence of the notations expressing the phonemes or phoneme groups every 10 ms. The sequence is again transformed into possible phonemic strings which are called input words henceforth and are convenient for the comparison with the contents of the dictionary. The Hammings distance between each input word and each item of the contents of the dictionary is computed where the notations of phonemes and phoneme groups are expressed by 9 bits binary vectors. The item in the dictionary nearest to one of the input word is selected as the output of the recognition system. The experiments were carried out with the utterance of five speakers from whose utterances the standard patterns for P1, P2 and Pe3 distribution had been made. The recognition score was 96% for the 20 city names involving all kinds of phonemes. The speech samples were increased to 166 city names and 82% of the utterances of three speakers were correctly recognized by adding the possible combination of phonemes to every word. Next, 13 different speakers uttered 51 city names having long distance between each other, the recognition score was found to be 94% when the speakers were permitted to repeat their utterances for three times.

Estimation of the waveform of a sound source by using an iterative technique with many sensors

This paper proposes a method to estimate the waveform of sound sources. This is accomplished by using a microphone array where the microphones are not situated close to the sound sources. The waveform is estimated using the sound velocity and positions of the microphones and sound sources as the only fixed parameters. The positions of the sound sources can be estimated in advance by using a conventional method such as multiple signal classification method (MUSIC). An iteration method is introduced to reduce the effect of direct sounds and of significant image sources due to reflected noise and/or other sound sources, whose positions are known. Computer simulations and an experiment with loudspeakers were conducted to demonstrate the effectiveness of our method.