Expressive Synthetic SpeechsitenavigationExpressive Synthetic Speech (pictures taken from P. Ekman)last update: Nov. 5th 2008This is a collection of examples of synthetic affective speech conveying an emotion or natural expression and maintained by FelixBurkhardt. Some of these samples are direct copies from natural data,others are generated by expert-rules or derived from data-bases. The emotional labels "anger", "fear", "joy" and"sad" are my (short) designators for "the big four" basic emotions,not neccessarily the authors' ones. Examples of German actorssimulating emotional arousal can be found here.Examples of German text-to-speech synthesizers can be found here.Please, feel encouraged to let me know about own or missingattempts to simulate emotional speech!(felixbur@gmx.de)contentscomparing simulation of anger, fear, joy and sadnessother simulationsrelated examplesfurther linksprojects concerning emotional speechchangelogAll Audiofiles are Mp3-format (64 or 32 kB/s)authorvisualaffil.year (approx)descriptionneutralangerjoysadfear N. Audibert, V. Aubergé, A. Rilliard ICP2006Copy prosody and intensity from satisfied and sad speech to neutralspeech using PSOLA technique. See the article "The Prosodic Dimensionsof Emotion in Speech: the Relative Weights of Parameters", Interspeech2005 (Lisbon), for details -- StephanBaldes DFKI1999Rule based emotion simulation with Entropic's formant TTS engine TrueTalk, based on Cahn's affect editor approach--MurtazaBulut, Carlos Busso, Serdar Yildirim, Abe Kazemzadeh, Chul Min Lee, Sungbok Lee, Shrikanth Narayanan USC/Sail2005Emotional voice-conversion by changing prosody(TD-PSOLA) and spectrum (LPC modification). The samplesdemonstrate neutral to target-emotion conversion. See the article"Investigating the role of phoneme-level modifications in emotionalspeech resynthesis", Proc Interspeech 2005 fordetails- MurtazaBulut, ShriNarayanan, Ann SyrdalUSC/Sail,AT&T2002Diphone synthesis done with hand-crafted diphonesand copy-prosody for the appropriate emotion. See the article "ExpressiveSpeech Synthesis Using a Concatenative Synthesizer", Proc. ICSLP 2002, fordetails-Felix Burkhardt T-Systems2005 emofilt: rule based simulation(prosody only) with MBROLA.German male voice (de6) (neutral prosody txt2pho)English male voice (en1) Frensh male voice (fr1)Greek male voice (gr2)Dutch male voice (nl2)Hungarian male voice (hu1)Italian male voice (it3)Turkish male voice (tr1)1998emofilt old version2000emoSyn 1: rulebased simulation with formant-synthesizer (Sensyn Version), the neutral sentence iscopy-synthesis, more examples here2000emoSyn 2: copysynthesis with formant-synthesizer (Iles & Simmons Version)orig.synth.1998esps2mbrola: prosody-copy synthesis with MBROLAorig.synth. Joao P. Cabral L2F INESC-ID Lisboa2005Transformation of neutral speech usingLP-PSOLA. Changes pitch, duration and energy as well asvoicequality (by transforming the residual). Emotion-rules werederived from literature. See the article "Pitch-SynchronousTime-Scaling for Prosodic and Voice Quality Transformations",Proc. Interspeech 2005 for details.male voice female voiceJanetCahn MIT1989 ?AffectEditor: rule based emotion simulation with DecTalk- Piero Cosi, FabioTesser, RobertoGretter, Carlo Drioli,Graziano Tisato ISTC-SPFD2004EmotiveMbrola: Italian concatenation synthesis with the Festival speechsynthesis framework and MBROLA voices.The prosody was learned from emotional database (CART). An articleappeared at the Interspeech 2005.male voice female voice with manipulation of voicequalitymale voice female voiceBjörn Granström,Rolf Carlson ? KTH1998 ?KTH RoyalInstitute of Technology (orig. link broken): swedish copyformant-synthesis.-authorvisualaffil.year (approx)descriptionneutralangerjoysadfear Akemi Iida ATR2000 ?ChatrEmotion: japanese concatenationsynthesis using emotional databases with CHATR. See the article "A Speech Synthesis System with Emotion for Assisting Communication", Proc. ISCA Workshop on Speech and Emotion, Belfast, 2000. Since than Chatrhas been expanded to NATR-- new CHATR with Emotion--Gregor HoferUniv. Edinburgh2004Gregor Hofer's master'sthesis, unit selection database recorded in neutral as well ashappy and angry style. Moresamples.--half-emotional by mixing units from neutral andemotional data-- IgnasiIriondo, Francesc Alías, Javier Melenchón, M. AngelesLlorca Univ. RamonLull2003Catalan diphone synthesis with emotion rules, seearticle "Modeling and Synthesizing Emotional Speech for CatalanText-to-Speech Synthesis" (Proc. ADS 2004) for details-Cynthia Breazeal MIT2000Kismet:rule based emotion simulation with DecTalk (like AffectEditor)with words Keisuke Miyanaga, Makoto Tachibana,Junichi Yamagishi, Koji Onishi, Takashi Masuko,TakaoKobayashi Tokyo Institute oftechnology, Kobayashi Lab.2004HMM (data-based) modeling of emotional expression(spectral and prosodic), enables mixing: see article HMM-BasedSpeech Synthesis with Various Speaking Styles Using ModelInterpolation or further demos:each emotion individually modeled- emotion as contextual factor (likephonetic/linguistic factors)-Juan M. MonteroMartínez Univ. Madrid1998 ?montero1: rule based emotion simulation with spanishdiphone-synthesizer-montero2: rule based emotion simulation withKTH-formant-synthesizer- Shinya Mori, Tsuyoshi Moriyama, Shinji Ozawa Dept.of Media and Image Technology, Tokyo Polytechnic University2008PSOLA transformation based on data-base trained prosody modificationrules. The algorithm includes possibility of graded expressions.See the article "Emotional Speech Synthesis UsingSubspace Constraints in Prosody", Proc. ICME 2006, or follow thislinkhalf-full- IainMurray Univ. Dundee1989 ?HAMLET:rule-based simulated emotions with formant speech synthesis 1997Laureate:hand optimized concatenation tts 1997LAERTES:rule-based simulation using laureatePierre-yves Oudeyer Sony2000cartoon speech:nonsense speech for sony pet-robots based on concatenativesynthesis. Article: Oudeyer P-Y. "The Synthesis of CartoonEmotional Speech", Proc. of the 1st International Conference onProsody, Aix-en-Provence, eds. B. Bel; I. Marlien. 2002low intensity-high intensity-with japanese words- ? IBM Tokyo ResearchLaboratory2000 ?ProTalker, fromIBM tokyo research laboratory.--Erhard Rank/Hannes Pirker ÖFAI1998VieCtoS:demos of the demisyllable LPC-synthesizer VieCtoS of the AustrianResearch Institute for Artificial Intelligence (ÖFAI) copyingemotional speech MarcSchröder DFKI2002While working on the NECA-Projectand his ph.d. thesis Marc developed a system capable ofproducing emotional speech based on a description from emotionaldimensions (arousal, valence, potency). I tried to map his resultsto basic emotions. The system is based on MBROLA as DSPand MARY as NLP. In order tocontrol voice-quality, six databases for MBROLA were developed: Formale and female each normal, tense voice and lax voice.-Jun Sato?1998 ?Japanese demo of emotional synthetic speechgenerated by art. neural nets. Article: J. Sato and S. Morishima,"Emotion modeling in speech production using emotion space", inProc. IEEE Int. Workshop on Robot and Human Communication, Tsukuba,Japan, Nov. 1996- Oytun Türk and Marc SchröderDFKI2008GMM based voice conversion, i.e. the spectral envelope of neutral speech gets transfered to the target emotional speech in order to simulate emotional voice quality without having to record a whole emotional database. See the article "A Comparison of Voice Conversion Methods for Transforming Voice Quality in Emotional Speech Synthesis", Proc. Interspeech 2008, Brisbane.orig. -synth. (GMM method)- E. Zovato,A. Pacchiotti, S. Quazza, S. Sandri Loquendo2004From Loquendo. Rule-based prosodyPSOLA-like manipulation of non-uniform unit-selection engine, seeTowards emotionalspeech synthesis: a rule based approach. Presented at the 5thISCA Workshop on Speech Synthesis in Pittsburgh 2004. other examples from Loqendo-examples not simulating the "big four"authorvisualaffil.year (approx)descriptionsamplesGreg Beller IRCAM2005Transform real voices thanks to a content-based transformation with a phase vocoder algorithm.Time-stretch and transpose coefficients change over the utterance depending on the expressivity and the context of units(whether they're part of an accentuated syllable or whether they'reconsonants, for instance...).originalsadtransformationboredtransformationfrightenedtransformation? Cepstral2004Non-uniform unit selection. Databases recorded with acertain style.Damian: dark personalityDuchess: sensitive voiceShouty: non-sensitive voice? ETI Eloquence (as Eloquent belongs to Scansoftnow, the link is broken)1998Rule-based formant-syntresis. The emotionalexpression was hand-optimized (demonstration from Eloquent)expressive maleexpressive femaleEllen Eide et al IBMWatson research Center2004Non-uniform unit-selection trained with anexpressive prosody model, paralinguistic events and expressiveunits. Described in the article A Corpus-Based Approach to<AHEM/> Expressive Speech Synthesisgood newsbad newsquestionotherfor comparison: IBM-CTTS taken from website. Notethat research-engine is advanced technology compared toproduct. Enrico Zovato et alLoquendo2004Non-uniform unit-selection enriched byparalinguistic events and expressive units. Other examples from Loquendo.German (Kathrin)/(Stefan)French (Juliette)English (Simon) /F. Malfrere MBROLIGNfrom TCTS Lab, Mons1999data-based prosody synthesis with MBROLA (thedecision-tree algorithm was trained on a database spoken with theappropriate affect)neutralnervousastonishedshy John Stallo Metafacefrom Curtin University2000Implementation of Virtual Human Markup Language, whichincludes emotional expression. TTS based on Festival andJohn Stallo's work on "Simulating Emotional Speechfor a Talking Head" which implements prosody rules.happy-cryhappy-go-luckyso-afraid? ModelTalkerfrom the University of Delaware1998biphone synthesis (biphone-inventory searches forbest diphones at synthesis time). The emotional expression wasgenerated by prosody-rulesneutralhappysurprisedfrustratedsadcontradictiveassertive Nick Campbell. further samples NATR from ATR2004Non-uniform unit-selection working on a very largedatabase (recording a woman for 3 years) including affectivelabeling and extralinguistic sounds. See the article"Extra-Semantic Protocols; Input Requirements for the Synthesis ofDialogue Speech" (Proc. ADS 2004) for details.sample of a telefone conversation between NATR(female voice) and a young man (originally talking to hisfriend). ? Rhetorical, now owned by Scansoft. Eduardo2002non-uniform unit-selection synthesis done by Rhetorical, characterfor a Jim Beam marketing campaign, manually optimizedin conversation with rhetorical's voice "americanvalley girl"MarcSchröderDFKI2002more examples (see above)excitedscaredsadangry 1angry 2angry 3boredcontent 1content 2content 3happyMARYDFKI2005Limited domain unit-selection with emotional units.excited soccer reportsMariet Theune HumanMedia Interaction lab from Univ. Twente2006Extracting prosodic rules to enhance speaking style for storytellers and apply to TTS. See the article "Generating Expressive Speech for StorytellingApplications" by Mariet Theune, Koen Meijs, Dirk Heylen and Roeland Ordelman. IEEE Transactions on Audio, Speech and Language Processing 14(4)no suspensesudden climaxincreasing climaxRaul Fernandez and Bhuvana Ramabhadran, IBM TJ Watson research centerEmphatic Speech2007Classifying emphatic speech in a non-uniform unit selection database. See the article "Automatic Exploration of Corpus-Specific Properties for Expressive Text-to-Speech." by Raul Fernandez and Bhuvana Ramabhadran. Proc. 6th ISCA workshop on speech synthesis, Bonn, 2007baseline neutral units with normal textbaseline plus collected emphatic units with marked emphasis textmined emphasis corpus with marked emphasis text? Talkback2000 (original link broken)1998concatenation synthesis, demonstration fromTalking Technologiepressive maleexpressive femalerelated examplesVocal Tract Lab articulatory synthesizer singing dona nobis pacemGlove Talk formant synthesis controlled by data glove, for further information see Sidney S. Fels and Geoffrey E. Hinton. Glove-TalkII: A neural network interface which maps gestures to parallel formant speech synthesizer controls. IEEE Transactions on Neural Networks. Volume 9. No. 1. Pages 205-212. 1998., see Sageev Oore demonstrate Glove Talk (43 Mb)Pavarobotti the singing robot: articulatory synthesis HAL From the film "2001: a space odyssey", the ideal(NOT a synthesizer but an actor!)contentsad/threateningregretfulFirst song insynthetic speech "Bicycle Built for Two", Bell Laboratories, by LouisGerstman and Max Mathews, 1961. This song was reprised by the `decorticated' Hal. ESPER Extracting Speaker Information From Children'sStories for Speech Synthesis: non-uniform unit selection Feelixfrom LolaCanamero and JacobFredslund Emotional music by Rasmus Lunding for a lego robot displayingemotions by mimic.missing samples (todo list ;-)Constructing emotional speech synthesizers with limited speechdatabase by Ryosuke Tsuzuki, Heiga Zen, Keiichi Tokuda, TadashiKitamura, Murtaza Bulut, Shrikanth S. NarayananSynthesis ofEmotional Speech Using Prosodically Balanced VCV Segments by Y.Niimi, M. Kasamatu, T. Nishimoto and M. Araki. Kyoto Institute ofTechnologyDFKI's Safira Affective Speech Module P. Gebhard, E. Andre.DFKI"a high quality emotional speech synthesis method", Sakamoto& Saito. IBM Japan"Corpus-Based Emotional Speech Synthesis Based on Generation ProcessModel and Its Evaluation", Sato, Hirose & Minematsu.University of TokyoGenerationof Emotions by a Morphing Technique in English, French andSpanish by P. Boula de Mareüil, P. Célérier, J.Toen. LIMSI-CNRS/Elanlinks about emotional speech-synthesisMIT'sKismetVirtual Human MarkupLanguage, Curtin UniversityEmotional SpeechHomepage of the DSPLab, university. of MariborGeneva EmotionResearch GroupAffectiveComputing, MIT, Media LabJean-MarcFellous and Eva Hudlicka: Emotion Home PageLolaCañamero's Emotion ForumChristophBartneck's Affective Computing Portalprojects that deal with affective speechnamedescriptiontimeframepartnersSpeech and EmotionWe study the effects of emotional state on speech, as well as the effects of emotional speech on listeners. This is achieved by recording skin conductance, goose bumps, blood pressure, and other peripheral measures of emotional state as well as vocal parameters. The outcome of these analyses is not only useful for basic research in emotion psychology, but also for clinical research and forensic studies, in the areas of developmental and pedagogical psychology, as well as in industrial and organizational psychology.10/2007-7/2008Univ. Kiel, LMU MunichHUMAINEHUMAINE (Human-MachineInteraction Network on Emotion) is a Network of Excellence in the EU'sSixth Framework Programme, in the IST (Information SocietyTechnologies) Thematic Priority. HUMAINE aims to lay thefoundations for European development of systems that can register,model and/or influence human emotional and emotion-related statesand processes - 'emotion-oriented systems'.2004-2008Queen's University, Belfast, DeutschesForschungszentrum für Künstliche Intelligenz GmbH,Institute of Communication and Computer Systems - NationalTechnical University of Athens, Université de Genève,FPSE, University of Hertfordshire, Istituto Trentino Di Cultura,Université de Paris VIII, ÖsterreichischeStudiengesellschaft für Kybernetik, Kungliga TekniskaHögskolan, Stockholm, Universität Augsburg,Università Degli Studi di Bari, Ecole PolytechniqueFédérale de Lausanne,Friedrich-Alexander-Universität , Università Degli Studidi Genova, University of Haifa, Imperial College of Science,Technology and Medicine, Inesc Id - Instituto de Engenharia deSistemas e Computadores: Investigação e Desenvolvimentoem Lisboa, King's College, London, Centre National De La RechercheScientifique, University of Oxford, University of Salford, Tel AvivUniversity, Trinity College, La Cantoche Production, FranceTélécom SA, T-Systems Nova GmbH,, Instituto SuperiorTécnico, Lisbon, University of Southern California, University of Zagreb, University of Twente, University of Sheffield, University of Leeds,Institute of Cognitive Sciences and Technologies, Rome. ErmisEmotionally Rich Man-machine IntelligentSystem, EU-project. Scope: The development of a prototypesystem for human computer interaction than can interpret its users'attitude or emotional state, e.g., activation/interest,boredom, andanger, in terms of their speech and/or their facial gestures andexpressions Adopted technologies: Linguistic and paralinguisticspeech analysis and robust speech recognition, facial expressionanalysis, interpretation of the user's emotional state usinghybrid, neurofuzzy, techniques, while being in accordance with theMPEG-4 standard.2002-2005ALTEC S.A., ILSP, ICCS-NTUA, EYETRONICS EYE, KULBelgium, Queens University Belfast, KCL UK, MIT Germany, FRANCETELECOM, BRITISH TELECOMJST/CRESTESP(Japan Science and Technology) / (Core Researchfor Evoluational Science and Technology) Expressive SpeechProcessing. The goal of the five-year ESP Project is to producea corpus of natural daily speech in order to design speechtechnology applications that are sensitive to the various ways inwhich people use changes in speaking style and voice quality tosignal the intentions underlying each utterance, i.e., to addinformation to spoken utterances beyond that carried by the text orthe words in the speech alone. The corpus is to include emotionalspeech, but also samples to illustrate attitudinal aspects ofspeech, such as politeness, hesitation, friendliness, anger, andsocial-distance. The most obvious applications of the resultingtechnology will be in speech synthesis, but the research alsoinvolves speech-recognition technology for the labeling andannotation of the speech databases, and the development of agrammar of spoken language in order to take into accountsupra-linguistic (i.e., paralinguistic and extralinguistic)information.2000-2005ATR, NAIST Graduate Institute, and KobeUniversity PF-StarPreparing future multisensorial interactionresearch, EU-project. Scope: PF-STAR intends to contribute toestablish future activities in the field of Multisensorial andMultilingual communication (Interface Technologies) on firmer basesby providing technological baselines, comparative evaluations, andassessment of prospects of core technologies, which future researchand development efforts can build from. To this end, the projectwill address three crucial areas: technologies for speech-to-speechtranslation, the detection and expressions of emotional states, andcore speech technologies for children. For each of them, promisingtechnologies/approaches will be selected, further developed andaligned towards common baselines. The results will be assessed andevaluated with respect to both their performances and futureprospects.2002-2004ITC - irst, Italy, RWTH Computer ScienceDepartment, Germany, Institute for Pattern Recognition ofFriedrich-Alexander-Universität Erlangen-Nürnberg,Germany, Interactive Systems Laboratories at UniversitaetKarlsruhe, Germany, Kungl Tekniska Högskolan, Sweden,Department of Electronic, Electrical & Computing Engineering ofthe University of Birmingham, United Kingdom, Istituto di Scienze eTecnologie della Cognizione, Sezione di Padova - "Fonetica eDialettologia", ItalyNecaEuropean Project: NECA promotes the concept ofmulti-modal communication with animated synthetic personalities. Aparticular focus in the project lies on communication betweenanimated characters that exhibit credible personality traits andaffective behavior. The key challenge of the project is thefruitful combination of different research strands includingsituation-based generation of natural language and speech,semiotics of non-verbal expression in situated socialcommunication, and the modelling of emotions and personality.2001-2003OFAI - Austrian Research Institute for ArtificialIntelligence, DFKI German Research Center for ArtificialIntelligence,FREESERVE UK internet portal, Information TechnologyResearch Institute at the University of Brighton, Sysis InteractiveSimulations AG ,Institute of Phonetics at the University of theSaarland InterfaceEuropean Project to make man/machine interactionmore natural. The objective of the project is to define new modelsand implement advanced tools for audio-video analysis, synthesisand representation in order to provide essential technologies forthe implementation of large-scale virtual and augmentedenvironments.2000-2002DIST - University of Genoa, L&H Belgium, ImageCoding Group - Linköping University Sweden, UniversitatPolitecnica de Catalunya, Ecole Polytechnique Fédérale deLausanne, Université de Genève, FPSE, Informatics andTelematics Institute Greece, Tecnologia Automazione Uomo,University of Maribor Slovenia, Curtin University of TechnologyAustralia, Umea University Sweden, Centre National de la RechercheScientifique France, W Interactive SA FraMimicVoice, Accent and Emotion Adaptive Text toSpeech Synthesis. The objective of this project is to develop aspeaker-adaptive text to speech synthesis with applications to highquality automatic voice dialogue, personalised voice for thedisabled, broadcast studio voice processing, interpreted telephony,very low bit rate phonetic speech coding, and multimediacommunication.2000 ?Brunel University, London TU-BerlinDFG-Projekt: Phonetische Reduktion undElaboration bei emotionaler Sprechweise. recording of Germanemotional databaseand analysis.1998-2000TU-BerlinEMOVOXVoice variability related to speaker-emotionalstate in Automatic Speaker Verification. Extending the Verivoxapproach. this project aims to systematically explore the effectsof transient speaker state changes on the acoustic speech signal,by plotting the space of speaker state-dependent variation inspeech. Based upon the knowledge gained by analyzing the speechrecorded from speakers in a number of induced cognitive andemotional states, new methods of structured training in ASV systemswill be developed. To collect emotional speech data for the Emovoxproject, we will create an interactive computer program designed toinduce various target emotions and stress in speakers.1999-2000 ?Université de Genève, FPSE SUSASSpeech Under Simulated and ActualStress.This database was put together by Duke University andAir Force Research Laboratory for researchers who are interested inthe characteristics and effects of speech under stress on speechprocessing recognizers. SUSAS was created at the Robust SpeechProcessing Laboratory in the Department of Electrical and ComputerEngineering at Duke University.1997 ?Univ. of Colorado at Boulder, CSLR, RSPLVeriVoxVoice Variability in Speaker Verification.The main aim of VeriVox is to improve the reliability of automaticspeaker verification (ASV), by developing novel,phonetically-informed methods for coping with the variation in aspeaker's voice.1996 ?Queen Margaret College, Edinburgh, IKP, RheinischeFriedrich-Wilhelms-Universität Bonn, Faculté dePsychologie et des Sciences de l'Education, University of Geneva,University of Cambridge, Trinity College, Dublin, University ofCambridge, Laboratoire Parole et Langage, CNRS Université deProvence, Ensigma,CNRS, Délégation RégionaleNormandie Univ. ofReadingthe Emotion in Speech Project. Of the manytypes of suprasegmental and affective information that have beenfound to occur in speech, relatively few have been coded in such away as to permit inclusion of them in large-scale machine-readablespeech databases. However, as the demand for more natural andunconstrained speech material grows, it becomes increasinglynecessary to look at ways of doing this. This project broughttogether expertise in phonetics and phonology and in cognitivepsychology in order to examine emotional speech and to produce adatabase of such speech to put alongside the emotionally neutralmaterial found in most spoken language databases.1995-2000 ?University of Reading, Department of Psychology atthe University of LeedsVAESSVoices, Attitudes and Emotions in SpeechSynthesis. The aim of the VAESS project is to develop a fullyportable (hand-held) communicator with versatile, high qualityspeech output. By combining the latest advances in speechtechnology withstate-of-the-art hardware, the capabilities ofcurrent speech prostheses will be extended. A deliberate choice wasmade to base the communicator around a standard personal computer.The potential uses of the portable communicator are then increasedto encompass those of any equivalent computer; there aresignificant benefits from this in the workplace where standardcomputer applications are needed. The VAESS Project is funded bythe European Union Technology Initiative for the Disabled andElderley Programme (TIDE).1995 ?Sheffield University, Center forPersonKommunikation at the University of Aalborg. Department ofSpeech and Music Acoustics at KTH in Stockholm, Telia PromoterInfovox AB in Stockholm, BiDesign Ltd in Tamworth., BarnsleyDistrict General Hospital NHS Trust VOX- 6298The Analysis and Synthesis of SpeakerCharacteristics. The VOX Working Group is investigating speechdatabases with different types of speakers, different affectiveconditions of emotion and attitude, and different casual versuscareful styles of speaking: each considered with reference toacoustic, perceptual and physiological representation. Speechsynthesis can be used to empirically test suchcharacterisations.1992-1995Université de Genève,IKP-Universität Bonn, CNRS-Institut de Phonetique, CNRS -LIMSI, Trinity College Dublin, KTH, University of Cambridge,University of Reading, University of SheffieldChangelogNov 5th 2008; added Moriyama samples.Oct 8th 2008; added Oytun Türk samples.February 5th 2008; added emofilt Dutch, Hungarian and Italian samples.Nov 2nd 2007; added Glove Talk and Vocal Tract Lab and IBM Fernandez samples.Oct 11th 2006; added Mariet Theune samples.June 19th 2006; added Audibert samples.February 17th 2006; added MARY unit selection samples.Decemer 14th 2005; added Beller samples.October 14th 2005; added Cepstral samples.October 10th 2005; added Sail voice conversion samples.September 19th 2005; added J.P. Cabral samples.August 29th 2005; added ESPER sample.August 9th 2005; added Metaface sample.August 3rd 2005; added emofilt Greek andTurkish samples.April 29th 2005; added Loquendoextralinguistics Stefan sample.April 25th 2005; added Loquendoextralinguistics samples.March 14th 2005; added Gregor Hofersamples.March 5th 2005; added ESP project.March 2nd 2005; added Kobayashisamples.February 16th 2005; added new emofiltsamples.February 15th 2005; added Piero Cosi samples and PF-Starproject.September 24th 2004; added more Oudeyer samplesSeptember 10th 2004; added Bulut/Narayanan/Syrdal samplesSeptember 2nd 2004; added Feelix sampleAugust 27th 2004; added Ignasi Iriondo and Nick CampbellsamplesAugust 5th 2004; added Stephan Baldes' samplesAugust 4th 2004; added Kismet (affect editor) samples withwordsAugust 2nd 2004; added Rhetorical's Eduardo sampleJuly 6th 2004; added IBM Watson Lab samplesJuly 4th 2004; added Loquendo samplesJune 23th 2004: added ProTalkerDemos ofTTS-Systems for Germancomments:felixbur@gmx.deband  |
|
