Articulatory Modelling 1
HomeFull List of Titles
1: ICSLP'98 Proceedings
Keynote SpeechesText-To-Speech Synthesis 1Spoken Language Models and Dialog 1Prosody and Emotion 1Hidden Markov Model Techniques 1Speaker and Language Recognition 1Multimodal Spoken Language Processing 1Isolated Word RecognitionRobust Speech Processing in Adverse Environments 1Spoken Language Models and Dialog 2Articulatory Modelling 1Talking to Infants, Pets and LoversRobust Speech Processing in Adverse Environments 2Spoken Language Models and Dialog 3Speech Coding 1Articulatory Modelling 2Prosody and Emotion 2Neural Networks, Fuzzy and Evolutionary Methods 1Utterance Verification and Word Spotting 1 / Speaker Adaptation 1Text-To-Speech Synthesis 2Spoken Language Models and Dialog 4Human Speech Perception 1Robust Speech Processing in Adverse Environments 3Speech and Hearing Disorders 1Prosody and Emotion 3Spoken Language Understanding Systems 1Signal Processing and Speech Analysis 1Spoken Language Generation and Translation 1Spoken Language Models and Dialog 5Segmentation, Labelling and Speech Corpora 1Multimodal Spoken Language Processing 2Prosody and Emotion 4Neural Networks, Fuzzy and Evolutionary Methods 2Large Vocabulary Continuous Speech Recognition 1Speaker and Language Recognition 2Signal Processing and Speech Analysis 2Prosody and Emotion 5Robust Speech Processing in Adverse Environments 4Segmentation, Labelling and Speech Corpora 2Speech Technology Applications and Human-Machine Interface 1Large Vocabulary Continuous Speech Recognition 2Text-To-Speech Synthesis 3Language Acquisition 1Acoustic Phonetics 1Speaker Adaptation 2Speech Coding 2Hidden Markov Model Techniques 2Multilingual Perception and Recognition 1Large Vocabulary Continuous Speech Recognition 3Articulatory Modelling 3Language Acquisition 2Speaker and Language Recognition 3Text-To-Speech Synthesis 4Spoken Language Understanding Systems 4Human Speech Perception 2Large Vocabulary Continuous Speech Recognition 4Spoken Language Understanding Systems 2Signal Processing and Speech Analysis 3Human Speech Perception 3Speaker Adaptation 3Spoken Language Understanding Systems 3Multimodal Spoken Language Processing 3Acoustic Phonetics 2Large Vocabulary Continuous Speech Recognition 5Speech Coding 3Language Acquisition 3 / Multilingual Perception and Recognition 2Segmentation, Labelling and Speech Corpora 3Text-To-Speech Synthesis 5Spoken Language Generation and Translation 2Human Speech Perception 4Robust Speech Processing in Adverse Environments 5Text-To-Speech Synthesis 6Speech Technology Applications and Human-Machine Interface 2Prosody and Emotion 6Hidden Markov Model Techniques 3Speech and Hearing Disorders 2 / Speech Processing for the Speech and Hearing Impaired 1Human Speech ProductionSegmentation, Labelling and Speech Corpora 4Speaker and Language Recognition 4Speech Technology Applications and Human-Machine Interface 3Utterance Verification and Word Spotting 2Large Vocabulary Continuous Speech Recognition 6Neural Networks, Fuzzy and Evolutionary Methods 3Speech Processing for the Speech-Impaired and Hearing-Impaired 2Prosody and Emotion 72: SST Student Day
SST Student Day - Poster Session 1SST Student Day - Poster Session 2Author Index
A B C D E F G H I
J K L M N O P Q R
S T U V W X Y Z
Multimedia Files
A Three-Dimensional Linear Articulatory Model Based on MRI Data
Authors:
Pierre Badin, Institut de la Communication Parlée, UPRESA CNRS 5009, INPG - Univ. Stendhal, Grenoble (France)
Gérard Bailly, Institut de la Communication Parlée, UPRESA CNRS 5009, INPG - Univ. Stendhal, Grenoble (France)
Monica Raybaudi, Unité INSERM U438, Grenoble (France)
Christoph Segebarth, Unité INSERM U438, Grenoble (France)
Page (NA) Paper number 14
Abstract:
Based on a set of 3D vocal tract images obtained by MRI, a 3D linear articulatory model has been built using guided Principal Component Analysis. It constitutes an extension to the lateral dimension of the mid-sagittal model previously developed from a radiofilm recorded on the same subject. The parameters of the 2D model have been found to be good predictors of the 3D shapes, for most configurations. A first evaluation of the model in terms of area functions and formants is presented.
SL980014.PDF (From Author)
SL980014.PDF (Rasterized)
TOP
On Loops and Articulatory Biomechanics
Authors:
Pascal Perrier, ICP - INPG & Université Stendhal - Grenoble (France)
Yohan Payan, TIMC - Université Joseph Fourier - Grenoble (France)
Joseph Perkell, Speech Group - RLE - MIT - Cambridge - MA (USA)
Frédéric Jolly, ICP - INPG & Université Stendhal - Grenoble (France)
Majid Zandipour, Speech Group - RLE - MIT - Cambridge - MA (USA)
Melanie Matthies, Speech Group - RLE - MIT - Cambridge - MA (USA)
Page (NA) Paper number 112
Abstract:
This study explores the following hypothesis: forward looping movements of the tongue that are observed in VCV sequences are due partly to the anatomical arrangement of the tongue muscles and how they are used to produce a velar closure. The study uses an anatomically based two-dimensional biomechanical tongue model. Tissue elastic properties are accounted for in finite-element modeling, and movement is controlled by constant-rate control parameter shifts. Tongue raising and lowering movements are produced by the model with the combined actions of the genioglossus, styloglossus and hyoglossus. Simulations of V1CV2 movements were made, where C is a velar consonant and V is [a], [i] or [u]. The resulting trajectories describe movements that begin to loop forward before consonant closure. Examination of subject data show similar looping movements. These observations support the idea that the biomechanical properties of the tongue could be the main factor responsible for the loops.
SL980112.PDF (From Author)
SL980112.PDF (Rasterized)
TOP
Magnetic Resonance Measurements of the Velum Port Opening
Authors:
Didier Demolin, Universite Libre de Bruxelles (Belgium)
Véronique Lecuit, Universite Libre de Bruxelles (Belgium)
Thierry Metens, Universite Libre de Bruxelles (Belgium)
Bruno Nazarian, Universite Libre de Bruxelles (Belgium)
Alain Soquet, Universite Libre de Bruxelles (Belgium)
Page (NA) Paper number 532
Abstract:
M.R.I. techniques have been used to describe velum opening of French vowels. Data based on 18 joined axial slices of 4 mm thickness were recorded with two subjects. Differences in velum opening are calculated from areas measured in the tract between the lowered velum and the back pharynx wall. A 3 D modelling of this tract is also proposed.
SL980532.PDF (From Author)
SL980532.PDF (Scanned)
0532_01.PDF(was: 0532_1.GIF) | MRI image File type: Image File Format: Image : GIF Tech. description: 514x510 pixels, 150 dpi, 113k Creating Application:: Photoshop Creating OS: MacOs 8.1 |
0532_02.PDF(was: 0532_2.GIF) | MRI image File type: Image File Format: Image : GIF Tech. description: 277x228 pixels, 72 dpi, 27k Creating Application:: Photoshop Creating OS: MacOs 8.1 |
0532_03.PDF(was: 0532_3.GIF) | 3-D reconstruction File type: Image File Format: Image : GIF Tech. description: 627x385 pixels, 72 dpi, 8 bits/pixel Creating Application:: Photoshop Creating OS: MacOs 8.1 |
TOP
Cantilever-Type Force-Sensor-Mounted Palatal Plate for Measuring Palatolingual Contact Stress and Pattern During Speech Phonation
Authors:
Masafumi Matsumura, Osaka Electro-Communication University (Japan)
Takuya Niikawa, Osaka Electro-Communication University (Japan)
Takao Tanabe, Osaka Electro-Communication University (Japan)
Takashi Tachimura, Osaka University (Japan)
Takeshi Wada, Osaka University (Japan)
Page (NA) Paper number 507
Abstract:
A 15-cantilever-type force-sensor unit is presented for the measurement of palatolingual contact stress and pattern during palatal consonant phonation. The force sensor unit is composed of a strain gauge and a cantilever, and is embedded in a thin palatal plate attached to the human hard palate. It is 3 mm wide, by 5 mm long, and 1.3 mm thick. The output of the force sensor unit at the low stress range of 0-64 kPa (0-5 gw) is proportional to the stress applied to the force sensing unit, with nearly no hysteresis. Measurement error of the force sensor is less than 1.7%. Error by mechanical interference among cantilever-type force sensors is less than 0.2%. The presented 15-cantilever-type force-sensor-mounted palatal plate allows for ready observation of the dynamic aspect of the palatolingual contact stress and patterns during the phonation of consonants.
SL980507.PDF (From Author)
SL980507.PDF (Rasterized)
TOP
Determination of the Vocal Tract Spectrum from the Articulatory Movements Based on the Search of an Articulatory-Acoustic Database
Authors:
Tokihiko Kaburagi, NTT Basic Research Laboratories (Japan)
Masaaki Honda, NTT Basic Research Laboratories (Japan)
Page (NA) Paper number 425
Abstract:
This paper presents a method for determining the vocal-tract spectrum from the positions of fixed points on the articulatory organs. The method is based on the search of a database comprised of pairs of articulatory and acoustic data representing the direct relationship between the articulator position and vocal-tract spectrum. To compile the database, the electro-magnetic articulograph (EMA) system is used to measure the movements of the jaw, lips, tongue, velum, and larynx simultaneously with speech waveforms. The spectrum estimation is accomplished by selecting database samples neighboring the input articulator position and interpolating the selected samples. In addition, phoneme categorization of the input position is performed to restrict the search area of the database to portions of the same phoneme category. Experiments show that the mean estimation error is 2.24 dB and the quality of speech synthesized from the estimated spectrum can be improved by using the phoneme categorization.
SL980425.PDF (From Author)
SL980425.PDF (Rasterized)
0425_01.WAV(was: 0425_01.WAV) | Speech samples synthesized with phoneme categorization are included
in the CD-ROM [SOUND 0425\_01.WAV] [SOUND 0425\_02.WAV] [SOUND 0425\_03.WAV]. File type: Sound File Format: Sound File: WAV Tech. description: None Creating Application:: Unknown Creating OS: Unknown |
0425_02.WAV(was: 0425_02.WAV) | Speech samples synthesized with phoneme categorization are included
in the CD-ROM [SOUND 0425\_01.WAV] [SOUND 0425\_02.WAV] [SOUND 0425\_03.WAV]. File type: Sound File Format: Sound File: WAV Tech. description: None Creating Application:: Unknown Creating OS: Unknown |
0425_03.WAV(was: 0425_03.WAV) | Speech samples synthesized with phoneme categorization are included
in the CD-ROM [SOUND 0425\_01.WAV] [SOUND 0425\_02.WAV] [SOUND 0425\_03.WAV]. File type: Sound File Format: Sound File: WAV Tech. description: None Creating Application:: Unknown Creating OS: Unknown |
TOP
An MRI Study On The Relationship Between Oral Cavity Shape And Larynx Position
Authors:
Kiyoshi Honda, ATR Human Information Processing Labs. (Japan)
Mark K. Tiede, ATR Human Information Processing Labs. (Japan)
Page (NA) Paper number 686
Abstract:
Individual variation of larynx position reflects human morphological differences and thus contributes to generating biological information in speech sounds. This study examines the factors of orofacial morphology that co-vary with larynx position based on MRI data collected for 12 Japanese and 12 English speakers. The materials are midsagittal craniofacial images, and the method is based on the measurement of angles and indices. Among all the measures examined, the aspect ratio of the oral cavity in the lateral view showed the highest correlation (r=0.87) with larynx height index (ratio of arytenoid - palatal plane distance and anterior nasal spine - nasopharyngeal wall distance), and a facial angle (angle of maxillary incisor - nasion - nasopharyngeal wall) showed the second highest correlation (r=0.66) with larynx height index. The result indicates that larynx position co-varies with oral cavity shape, being higher when oral cavity shape is flatter and more prognathic.