J. Sirigos

University of Patras

Author Statistics

Papers

Citation

H-Index

i-10 index

Research Trends

Author Order

Document Type

Co-Authors

Nikos Fakotakis

University of Patras

George Kokkinakis

University of Patras

G. Kokkinakis

University of Patras

N. Antoniades

Limassol General Hospital

D.A. Valkaniotis

University of Patras

Vassilis Darsinos

University of Patras

V. Darsinos

University of Patras

Cooperative Institutions

University of Patras

National and Kapodistrian University of Athens

Hella (Germany)

Limassol General Hospital

Zoological Research Museum Alexander Koenig

Bellvitge University Hospital

Centro de Investigación Biomédica en Red

Institut d'Investigació Biomédica de Bellvitge

University of Bucharest

Laboratoire d'Informatique pour la Mécanique et les Sciences de l'Ingénieur

Author Statistics

Papers

Citation

H-Index

i-10 index

Research Field

Accurate estimation of sinusoidal parameters in an harmonic+noise model for speech synthesis

Nikos Fakotakis J. Sirigos George Kokkinakis

Harmonic

Harmonic Analysis

10.21437/eurospeech.1999-252

Cite

Citations (5)

Text-independent off-line writer recognition using neural networks

D.A. Valkaniotis J. Sirigos N. Antoniades Nikos Fakotakis

A system of writer recognition using neural networks is described in this paper. The system is text independent and can be used for both identification and verification purposes. It consists of 20 multi-layer perceptrons as many, as the population of writers of the test. The letters used for training and testing were part of the Greek alphabet and were non-correlated. The system was tested on a total number of 5000 letters coming from the 20 writers. Error rates as low as 0.5% were achieved on test sets with more than 30 letters per set, in identification testing. In the verification testing the mean error was 1.2% on test sets with more than 15 letters per set. The response delay of the system was negligible (0.4 seconds on a conventional PC).

Perceptron

Identification

Line (geometry)

10.1109/icecs.1996.584456

Cite

Citations (1)

High performance text-independent speaker recognition system based on voiced/unvoiced segmentation and multiple neural nets

Nikos Fakotakis J. Sirigos George Kokkinakis

This paper presents a text-independent speaker recognition system based on the voiced segments of the speech signal. The proposed system uses feedforward MLP classification with only a limited amount of training and testing data and gives a comparatively high accuracy. The techniques employed are: the Rasta-PLP speech analysis for parameter estimation, a feedforward MLP for voiced/unvoiced segmentation and a large number (equal to the number of speakers) of simple MLPs for the classification procedure. The system has been trained and tested using TIMIT and NTIMIT databases. The verification experiments presented a high accuracy rate: above 99% for clean speech (TIMIT) and 74.7%, for noisy speech (NTIMIT). Additional experiments were performed comparing the proposed approach of using voiced segments with only vowels and all phonetic categories with results favorable to the use of voiced segments.

TIMIT

Speech segmentation

Feed forward

Feedforward neural network

10.21437/eurospeech.1999-239

Cite

Citations (2)

A high performance text independent speaker recognition system based on vowel spotting and neural nets

Nikos Fakotakis J. Sirigos

We present a text independent speaker recognition system based on vowel spotting and feed forward multilayer perceptrons (MLPs). The perceptual linear predictive (PLP) speech analysis technique was used for parameter estimation, a feed forward MLP for vowel spotting and a simple MLP for the classification procedure. To train and test the system we used the TIMIT database. We conclude with results of the speaker verification and identification process, showing that the system described has a high recognition accuracy (/spl sim/98%) using short test utterances (2.5 sec). It also has a real-time response, is easily adapted to new speakers and requires a small amount of data for training purposes (three sentences per speaker).

Spotting

TIMIT

Perceptron

Speaker identification

Multilayer perceptron

10.1109/icassp.1996.543207

Cite

Citations (12)

A high-performance vowel spotting system based on a multistage architecture

European Signal Processing Conference (1998)

J. Sirigos Nikos Fakotakis G. Kokkinakis

In this paper we present a novel multi-level vowel detection system with improved accuracy. Multi layer perceptrons (MLP), Discrete Hidden Markov Models (DSHMM) and heuristic rules are combined in three different levels to reduce the probability of false acceptance and rejection of vowel sounds. The TIMIT database was used to train and test this system. The rules are variable and are automatically customized by statistics extracted from the database, which concern the duration, the energy and the distance between vowels. The proposed method can easily be extended to languages other than English as long as a proper database exists for training the system. Its accuracy was measured to 99.22% using all the test data sets of the TIMIT database. Thus, the proposed vowel detection process can be reliably used for speech processing applications (speaker or speech recognition) where accurate vowel spotting algorithms are necessary.

TIMIT

Spotting

Perceptron

10.5281/zenodo.36502

Cite

Citations (2)

Vowel-non vowel decision using neural networks and rules

J. Sirigos V. Darsinos Nikos Fakotakis G. Kokkinakis

This paper describes a speaker independent vowel/non-vowel classifier based on neural networks and several rules. RASTA-PLP analysis of the speech signal resulting to mel-cepstral coefficients and a formant tracking method are used in order to provide the feature vectors for the MLP. To train and test the system we used a part of the TIMIT database. The results indicate that the performance of this classifier for speaker independent vowel classification is approximately 98.5% so it can be favorably used for speaker recognition or speech labeling purposes.

TIMIT

Mel-frequency cepstrum

Cepstrum

10.1109/icecs.1996.582917

Cite

Citations (2)

TEXT-INDEPENDENT OFF-LINE WRITER RECOGNITION USING NEURAL NETWORKS

Zenodo (CERN European Organization for Nuclear Research) (1996)

D.A. Valkaniotis J. Sirigos N. Antoniades Nikos Fakotakis

Line (geometry)

Text recognition

10.5281/zenodo.36306

Cite

Citations (2)

A hybrid syllable recognition system based on vowel spotting

Speech Communication (2002)

J. Sirigos Nikos Fakotakis George Kokkinakis

TIMIT

Spotting

10.1016/s0167-6393(02)00012-2

Cite

Citations (12)

Vowel-non vowel classification of speech using an MLP and rules

European Signal Processing Conference (1996)

J. Sirigos Vassilis Darsinos Nikos Fakotakis George Kokkinakis

In this paper we present a high precision speaker independent vowel/non vowel classifier based on a simple feed forward MLP (Multi Layer Perceptron) and several rules. RASTA-PLP analysis of the speech signal resulting to mel-cepstral coefficients and a formant tracking method are used in order to provide the feature vectors for the MLP. To train and test the system we used a part of the TIMIT database. The results indicate that the performance of this classifier for speaker independent vowel classification is approximately 97.25% so it can be favorably used for speaker recognition or speech labeling purposes.

TIMIT

Mel-frequency cepstrum

Perceptron

Multilayer perceptron

10.5281/zenodo.36335

Cite

Citations (3)

A hybrid ANN/HMM syllable recognition module based on vowel spotting

J. Sirigos Nikos Fakotakis George Kokkinakis

This paper presents a hybrid ANN/HMM syllable recognition module based on vowel spotting. An advanced multi-level vowel spotting method is used to achieve minimum vowel loss and accurate detection of the vowel location and duration. Discrete Hidden Markov Models (DSHMM), Multi Layer Perceptrons (MLP) and Heuristics (HR) are used for this purpose. A hybrid ANN/HMM technique is then used to recognize the syllables between the detected vowels. We replace the usual DSHMM probability parameters with combined neural network outputs. For this purpose both context dependent (CD) and context independent (CI) neural networks are used. Global normalization is employed on the parameters as opposed to the local normalization used on parameters in standard HMMs. Also, all parameters are estimated simultaneously according to the discriminative conditional maximum likelihood (CML) criterion. The tests were performed on the TIMIT and NTIMIT databases and showed significant performance improvement compared to similar systems.

TIMIT

Spotting

Discriminative model

Normalization

Keyword spotting

Perceptron

10.21437/eurospeech.1999-282x

Cite

Citations (0)