Keiichi TOKUDA


A Bayesian Approach to Image Recognition Based on Separable Lattice Hidden Markov Models
Kei SAWADA Akira TAMAMORI Kei HASHIMOTO Yoshihiko NANKAKU Keiichi TOKUDA 
Publication:   IEICE TRANSACTIONS on Information and Systems
Publication Date: 2016/12/01
Vol. E99-D  No. 12  pp. 3119-3131
Type of Manuscript:  PAPER
Category: Pattern Recognition
Keyword: 
image recognitionhidden Markov modelsseparable lattice hidden Markov modelsBayesian approachdeterministic annealing
 Summary | Full Text:PDF(972.5KB)

Image Recognition Based on Separable Lattice Trajectory 2-D HMMs
Akira TAMAMORI Yoshihiko NANKAKU Keiichi TOKUDA 
Publication:   IEICE TRANSACTIONS on Information and Systems
Publication Date: 2014/07/01
Vol. E97-D  No. 7  pp. 1842-1854
Type of Manuscript:  PAPER
Category: Pattern Recognition
Keyword: 
image recognitionhidden Markov modelsseparable lattice 2-D HMMstrajectory HMMs
 Summary | Full Text:PDF(1.6MB)

Integration of Spectral Feature Extraction and Modeling for HMM-Based Speech Synthesis
Kazuhiro NAKAMURA Kei HASHIMOTO Yoshihiko NANKAKU Keiichi TOKUDA 
Publication:   IEICE TRANSACTIONS on Information and Systems
Publication Date: 2014/06/01
Vol. E97-D  No. 6  pp. 1438-1448
Type of Manuscript:  Special Section PAPER (Special Section on Advances in Modeling for Real-world Speech Information Processing and its Application)
Category: HMM-based Speech Synthesis
Keyword: 
integrative modelHMM-based speech synthesisacoustic modelingmel-cepstral analysistrajectory HMM
 Summary | Full Text:PDF(1.7MB)

A Bayesian Framework Using Multiple Model Structures for Speech Recognition
Sayaka SHIOTA Kei HASHIMOTO Yoshihiko NANKAKU Keiichi TOKUDA 
Publication:   IEICE TRANSACTIONS on Information and Systems
Publication Date: 2013/04/01
Vol. E96-D  No. 4  pp. 939-948
Type of Manuscript:  PAPER
Category: Speech and Hearing
Keyword: 
speech recognitionacoustic modelingBayesian approachmodel structure integrationdeterministic annealing
 Summary | Full Text:PDF(548.8KB)

An Extension of Separable Lattice 2-D HMMs for Rotational Data Variations
Akira TAMAMORI Yoshihiko NANKAKU Keiichi TOKUDA 
Publication:   IEICE TRANSACTIONS on Information and Systems
Publication Date: 2012/08/01
Vol. E95-D  No. 8  pp. 2074-2083
Type of Manuscript:  PAPER
Category: Pattern Recognition
Keyword: 
image recognitionhidden Markov modelsvariational method
 Summary | Full Text:PDF(1.3MB)

Bayesian Context Clustering Using Cross Validation for Speech Recognition
Kei HASHIMOTO Heiga ZEN Yoshihiko NANKAKU Akinobu LEE Keiichi TOKUDA 
Publication:   IEICE TRANSACTIONS on Information and Systems
Publication Date: 2011/03/01
Vol. E94-D  No. 3  pp. 668-678
Type of Manuscript:  PAPER
Category: Speech and Hearing
Keyword: 
Bayesian approachspeech recognitionHMMcontext clusteringcross validation
 Summary | Full Text:PDF(453.4KB)

A Covariance-Tying Technique for HMM-Based Speech Synthesis
Keiichiro OURA Heiga ZEN Yoshihiko NANKAKU Akinobu LEE Keiichi TOKUDA 
Publication:   IEICE TRANSACTIONS on Information and Systems
Publication Date: 2010/03/01
Vol. E93-D  No. 3  pp. 595-601
Type of Manuscript:  PAPER
Category: Speech and Hearing
Keyword: 
HMMspeech synthesisdecision treecontext-clusteringMDL criterionembedded device
 Summary | Full Text:PDF(662.9KB)

A Reordering Model Using a Source-Side Parse-Tree for Statistical Machine Translation
Kei HASHIMOTO Hirofumi YAMAMOTO Hideo OKUMA Eiichiro SUMITA Keiichi TOKUDA 
Publication:   IEICE TRANSACTIONS on Information and Systems
Publication Date: 2009/12/01
Vol. E92-D  No. 12  pp. 2386-2393
Type of Manuscript:  Special Section PAPER (Special Section on Natural Language Processing and its Applications)
Category: Machine Translation
Keyword: 
phrase-based statistical machine translationreordering modelparse-treesyntactic information
 Summary | Full Text:PDF(254.4KB)

A Fully Consistent Hidden Semi-Markov Model-Based Speech Recognition System
Keiichiro OURA Heiga ZEN Yoshihiko NANKAKU Akinobu LEE Keiichi TOKUDA 
Publication:   IEICE TRANSACTIONS on Information and Systems
Publication Date: 2008/11/01
Vol. E91-D  No. 11  pp. 2693-2700
Type of Manuscript:  PAPER
Category: Speech and Hearing
Keyword: 
speech recognitionhidden Markov modelhidden semi-Markov modelweighted finite-state transducer
 Summary | Full Text:PDF(531.3KB)

The Nitech-NAIST HMM-Based Speech Synthesis System for the Blizzard Challenge 2006
Heiga ZEN Tomoki TODA Keiichi TOKUDA 
Publication:   IEICE TRANSACTIONS on Information and Systems
Publication Date: 2008/06/01
Vol. E91-D  No. 6  pp. 1764-1773
Type of Manuscript:  PAPER
Category: Speech and Hearing
Keyword: 
statistical parametric speech synthesisBlizzard Challenge 2006MGC-LSPMLLTfull covariance GV pdf
 Summary | Full Text:PDF(1.4MB)

A Hidden Semi-Markov Model-Based Speech Synthesis System
Heiga ZEN Keiichi TOKUDA Takashi MASUKO Takao KOBAYASIH Tadashi KITAMURA 
Publication:   IEICE TRANSACTIONS on Information and Systems
Publication Date: 2007/05/01
Vol. E90-D  No. 5  pp. 825-834
Type of Manuscript:  PAPER
Category: Speech and Hearing
Keyword: 
hidden Markov modelhidden semi-Markov modelHMM-based speech synthesis
 Summary | Full Text:PDF(452.5KB)

A Speech Parameter Generation Algorithm Considering Global Variance for HMM-Based Speech Synthesis
Tomoki TODA Keiichi TOKUDA 
Publication:   IEICE TRANSACTIONS on Information and Systems
Publication Date: 2007/05/01
Vol. E90-D  No. 5  pp. 816-824
Type of Manuscript:  PAPER
Category: Speech and Hearing
Keyword: 
HMM-based speech synthesisspeech parameter generationmaximum likelihood criterionover-smoothing effectglobal variance
 Summary | Full Text:PDF(1.2MB)

State Duration Modeling for HMM-Based Speech Synthesis
Heiga ZEN Takashi MASUKO Keiichi TOKUDA Takayoshi YOSHIMURA Takao KOBAYASIH Tadashi KITAMURA 
Publication:   IEICE TRANSACTIONS on Information and Systems
Publication Date: 2007/03/01
Vol. E90-D  No. 3  pp. 692-693
Type of Manuscript:  LETTER
Category: Speech and Hearing
Keyword: 
duration modelingspeech synthesishidden Markov model
 Summary | Full Text:PDF(54.1KB)

Details of the Nitech HMM-Based Speech Synthesis System for the Blizzard Challenge 2005
Heiga ZEN Tomoki TODA Masaru NAKAMURA Keiichi TOKUDA 
Publication:   IEICE TRANSACTIONS on Information and Systems
Publication Date: 2007/01/01
Vol. E90-D  No. 1  pp. 325-333
Type of Manuscript:  PAPER
Category: Speech and Hearing
Keyword: 
HMM-based speech synthesisBlizzard Challenge 2005STRAIGHTHSMMGV
 Summary | Full Text:PDF(700KB)

Continuous Speech Recognition Based on General Factor Dependent Acoustic Models
Hiroyuki SUZUKI Heiga ZEN Yoshihiko NANKAKU Chiyomi MIYAJIMA Keiichi TOKUDA Tadashi KITAMURA 
Publication:   IEICE TRANSACTIONS on Information and Systems
Publication Date: 2005/03/01
Vol. E88-D  No. 3  pp. 410-417
Type of Manuscript:  Special Section PAPER (Special Section on Corpus-Based Speech Technologies)
Category: Feature Extraction and Acoustic Medelings
Keyword: 
continuous speech recognitiontriphone HMMscontext clusteringBayesian networksvoice characteristicnoise environment
 Summary | Full Text:PDF(487.4KB)

Deterministic Annealing EM Algorithm in Acoustic Modeling for Speaker and Speech Recognition
Yohei ITAYA Heiga ZEN Yoshihiko NANKAKU Chiyomi MIYAJIMA Keiichi TOKUDA Tadashi KITAMURA 
Publication:   IEICE TRANSACTIONS on Information and Systems
Publication Date: 2005/03/01
Vol. E88-D  No. 3  pp. 425-431
Type of Manuscript:  Special Section PAPER (Special Section on Corpus-Based Speech Technologies)
Category: Feature Extraction and Acoustic Medelings
Keyword: 
DAEM algorithmacoustic modelingEM algorithmGMMsHMMs
 Summary | Full Text:PDF(680.4KB)

Applying Sparse KPCA for Feature Extraction in Speech Recognition
Amaro LIMA Heiga ZEN Yoshihiko NANKAKU Keiichi TOKUDA Tadashi KITAMURA Fernando G. RESENDE 
Publication:   IEICE TRANSACTIONS on Information and Systems
Publication Date: 2005/03/01
Vol. E88-D  No. 3  pp. 401-409
Type of Manuscript:  Special Section PAPER (Special Section on Corpus-Based Speech Technologies)
Category: Feature Extraction and Acoustic Medelings
Keyword: 
kernelsparsityprincipal component analysisfeature extractionspeech recognition
 Summary | Full Text:PDF(582.4KB)

Parameter Sharing in Mixture of Factor Analyzers for Speaker Identification
Hiroyoshi YAMAMOTO Yoshihiko NANKAKU Chiyomi MIYAJIMA Keiichi TOKUDA Tadashi KITAMURA 
Publication:   IEICE TRANSACTIONS on Information and Systems
Publication Date: 2005/03/01
Vol. E88-D  No. 3  pp. 418-424
Type of Manuscript:  Special Section PAPER (Special Section on Corpus-Based Speech Technologies)
Category: Feature Extraction and Acoustic Medelings
Keyword: 
speaker identificationGMMmixture of factor analyzersparameter sharingminimum classification error training
 Summary | Full Text:PDF(545.3KB)

On the Use of Kernel PCA for Feature Extraction in Speech Recognition
Amaro LIMA Heiga ZEN Yoshihiko NANKAKU Chiyomi MIYAJIMA Keiichi TOKUDA Tadashi KITAMURA 
Publication:   IEICE TRANSACTIONS on Information and Systems
Publication Date: 2004/12/01
Vol. E87-D  No. 12  pp. 2802-2811
Type of Manuscript:  PAPER
Category: Speech and Hearing
Keyword: 
kernelfeature spaceprincipal component analysisfeature extractionspeech recognition
 Summary | Full Text:PDF(425.6KB)

Mixture Density Models Based on Mel-Cepstral Representation of Gaussian Process
Toru TAKAHASHI Keiichi TOKUDA Takao KOBAYASHI Tadashi KITAMURA 
Publication:   IEICE TRANSACTIONS on Fundamentals of Electronics, Communications and Computer Sciences
Publication Date: 2003/08/01
Vol. E86-A  No. 8  pp. 1971-1978
Type of Manuscript:  Special Section PAPER (Special Section on Digital Signal Processing)
Category: 
Keyword: 
Gaussian mixture modelstatistical frameworkmel-cepstrumEM algorithm
 Summary | Full Text:PDF(1MB)

A Training Method of Average Voice Model for HMM-Based Speech Synthesis
Junichi YAMAGISHI Masatsune TAMURA Takashi MASUKO Keiichi TOKUDA Takao KOBAYASHI 
Publication:   IEICE TRANSACTIONS on Fundamentals of Electronics, Communications and Computer Sciences
Publication Date: 2003/08/01
Vol. E86-A  No. 8  pp. 1956-1963
Type of Manuscript:  Special Section PAPER (Special Section on Digital Signal Processing)
Category: 
Keyword: 
decision treecontext clusteringspeaker adaptationspeaker adaptive trainingaverage voice modelHMM-based speech synthesis
 Summary | Full Text:PDF(599.3KB)

A Context Clustering Technique for Average Voice Models
Junichi YAMAGISHI Masatsune TAMURA Takashi MASUKO Keiichi TOKUDA Takao KOBAYASHI 
Publication:   IEICE TRANSACTIONS on Information and Systems
Publication Date: 2003/03/01
Vol. E86-D  No. 3  pp. 534-542
Type of Manuscript:  Special Section PAPER (Special Issue on Speech Information Processing)
Category: Speech Synthesis and Prosody
Keyword: 
decision treecontext clusteringaverage voice modelHMM-based speech synthesisspeaker independent model
 Summary | Full Text:PDF(752.9KB)

Multi-Space Probability Distribution HMM
Keiichi TOKUDA Takashi MASUKO Noboru MIYAZAKI Takao KOBAYASHI 
Publication:   IEICE TRANSACTIONS on Information and Systems
Publication Date: 2002/03/01
Vol. E85-D  No. 3  pp. 455-464
Type of Manuscript:  INVITED PAPER (Special Issue on the 2000 IEICE Excellent Paper Award)
Category: Pattern Recognition
Keyword: 
hidden Markov modeltext-to-speechF0multi-space probability distribution
 Summary | Full Text:PDF(624.5KB)

Vector Quantization of Speech Spectral Parameters Using Statistics of Static and Dynamic Features
Kazuhito KOISHIDA Keiichi TOKUDA Takashi MASUKO Takao KOBAYASHI 
Publication:   IEICE TRANSACTIONS on Information and Systems
Publication Date: 2001/10/01
Vol. E84-D  No. 10  pp. 1427-1434
Type of Manuscript:  PAPER
Category: Speech and Hearing
Keyword: 
vector quantizationdynamic featuresspectral quantizationLSP parameters
 Summary | Full Text:PDF(640.5KB)

Text-Independent Speaker Identification Using Gaussian Mixture Models Based on Multi-Space Probability Distribution
Chiyomi MIYAJIMA Yosuke HATTORI Keiichi TOKUDA Takashi MASUKO Takao KOBAYASHI Tadashi KITAMURA 
Publication:   IEICE TRANSACTIONS on Information and Systems
Publication Date: 2001/07/01
Vol. E84-D  No. 7  pp. 847-855
Type of Manuscript:  Special Section PAPER (Special Issue on Biometric Person Authentication)
Category: 
Keyword: 
speaker identificationpitchmulti-space probability distributionGaussian mixture model (GMM)minimum classification error
 Summary | Full Text:PDF(921.7KB)

Lip Location Normalized Training for Visual Speech Recognition
Oscar VANEGAS Keiichi TOKUDA Tadashi KITAMURA 
Publication:   IEICE TRANSACTIONS on Information and Systems
Publication Date: 2000/11/25
Vol. E83-D  No. 11  pp. 1969-1977
Type of Manuscript:  PAPER
Category: Speech and Hearing
Keyword: 
hidden Markov modellip location normalizationlipreadingTulips1M2VTS
 Summary | Full Text:PDF(2.4MB)

A 16 kb/s Wideband CELP-Based Speech Coder Using Mel-Generalized Cepstral Analysis
Kazuhito KOISHIDA Gou HIRABAYASHI Keiichi TOKUDA Takao KOBAYASHI 
Publication:   IEICE TRANSACTIONS on Information and Systems
Publication Date: 2000/04/25
Vol. E83-D  No. 4  pp. 876-883
Type of Manuscript:  PAPER
Category: Speech and Hearing
Keyword: 
wideband speech codingmel-generalized cepstral analysismel-generalized cepstrumCELP coding
 Summary | Full Text:PDF(693.7KB)

LMS-Based Algorithms with Multi-Band Decomposition of the Estimation Error Applied to System Identification
Fernando Gil V. RESENDE,Jr Paulo S.R. DINIZ Keiichi TOKUDA Mineo KANEKO Akinori NISHIHARA 
Publication:   IEICE TRANSACTIONS on Fundamentals of Electronics, Communications and Computer Sciences
Publication Date: 1997/08/25
Vol. E80-A  No. 8  pp. 1376-1383
Type of Manuscript:  Special Section PAPER (Special Section on Digital Signal Processing)
Category: 
Keyword: 
adaptive signal processingfixed and variable step-size LMS algorithmsfilter bankssystem identification
 Summary | Full Text:PDF(629.5KB)

Multi-Band Decomposition of the Linear Prediction Error Applied to Adaptive AR Spectral Estimation
Fernando Gil V. RESENDE Jr. Keiichi TOKUDA Mineo KANEKO Akinori NISHIHARA 
Publication:   IEICE TRANSACTIONS on Fundamentals of Electronics, Communications and Computer Sciences
Publication Date: 1997/02/25
Vol. E80-A  No. 2  pp. 365-376
Type of Manuscript:  PAPER
Category: Digital Signal Processing
Keyword: 
digital signal processingAR spectral estimationfilter banksadaptive filteringrecursive least-squares algorithms
 Summary | Full Text:PDF(902.1KB)

Adaptive AR Spectral Estimation Based on Wavelet Decomposition of the Linear Prediction Error
Fernando Gil V. RESENDE Jr. Keiichi TOKUDA Mineo KANEKO 
Publication:   IEICE TRANSACTIONS on Fundamentals of Electronics, Communications and Computer Sciences
Publication Date: 1996/05/25
Vol. E79-A  No. 5  pp. 665-673
Type of Manuscript:  PAPER
Category: Digital Signal Processing
Keyword: 
digital signal processingspectral etimationwavelet theoryadaptive fulteringrecursive least-squares algorithms
 Summary | Full Text:PDF(725.7KB)

Time Series Analysis Based on Exponential Model Excited by t-Distribution Process and Its Algorithm
Junibakti SANUBARI Keiichi TOKUDA Mahoki ONODA 
Publication:   IEICE TRANSACTIONS on Fundamentals of Electronics, Communications and Computer Sciences
Publication Date: 1993/05/25
Vol. E76-A  No. 5  pp. 808-819
Type of Manuscript:  PAPER
Category: Digital Signal Processing
Keyword: 
exponential modelt-distributionM-estimatetime series analysisfinite-length cepstrum
 Summary | Full Text:PDF(788.1KB)

Speech Analysis Based on AR Model Driven by t-Distribution Process
Junibakti SANUBARI Keiichi TOKUDA Mahoki ONODA 
Publication:   IEICE TRANSACTIONS on Fundamentals of Electronics, Communications and Computer Sciences
Publication Date: 1992/09/25
Vol. E75-A  No. 9  pp. 1159-1169
Type of Manuscript:  PAPER
Category: Speech
Keyword: 
AR modelingt-distributionM-estimatespeech analysis
 Summary | Full Text:PDF(749.9KB)

2-D LMA Filters--Design of Stable Two-Dimensional Digital Filters with Arbitrary Magnitude Function--
Takao KOBAYASHI Kazuyoshi FUKUSHI Keiichi TOKUDA Satoshi IMAI 
Publication:   IEICE TRANSACTIONS on Fundamentals of Electronics, Communications and Computer Sciences
Publication Date: 1992/02/25
Vol. E75-A  No. 2  pp. 240-246
Type of Manuscript:  PAPER
Category: Digital Image Processing
Keyword: 
2-D digital filterLMA filter2-D spectral factorization
 Summary | Full Text:PDF(507.3KB)