Text-To-Speech Synthesis 2
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
Prosody Prediction for Speech Synthesis using Transformational Rule-based Learning
Authors:
Cameron S. Fordyce, Lernout & Hauspie Speech Products (USA)
Mari Ostendorf, Dept. Of ECE, Boston University (USA)
Page (NA) Paper number 682
Abstract:
Speech generation systems can benefit from the prediction of abstract prosodic labels from text input. Earlier methods of prosodic label prediction have relied on hand-written rules or on statistical methods such as decision trees. Statistical methods have the advantage of being automatically trainable and are portable to new domains. This research presents a new method for automatically training an abstract prosodic label predictor, transformational rule-based learning. This method is automatically trainable. Results will be presented for pitch accent location and phrase boundary prediction.
SL980682.PDF (From Author)
SL980682.PDF (Rasterized)
TOP
Representing the Environments for Phonological Processes in an Accent-Independent Lexicon for Synthesis of English
Authors:
Susan Fitt, Centre for Speech Technology Research, University of Edinburgh (U.K.)
Stephen Isard, Centre for Speech Technology Research, University of Edinburgh (U.K.)
Page (NA) Paper number 850
Abstract:
This paper reports on work developing an accent-independent lexicon for use in synthesising speech in English. Developing a lexicon for a new accent is a long process, and one potential solution to this problem involves the encoding of regional variation by means of keywords; so, rather than transcribing different phonemes for 'pool' in RP and in Scottish accents, we can simply say that the word contains the same vowel as in the keyword GOOSE. However, there are a number of theoretical and practical issues, which are discussed here. It is proposed that phonemic variation within accents be encoded in the lexicon by use of keyword symbols, while allophonic differences be derived by accent-specific rules. If we wish to include some stylistic variation this makes the lexicon more comprehensive but more complex. Finally, it is noted that even in keyword synthesis exception lists cannot be avoided.
SL980850.PDF (From Author)
SL980850.PDF (Rasterized)
TOP
Efficient Lexical Retrieval for English Text-to-Speech Synthesis
Authors:
Daniel Faulkner, Aculab PLC (U.K.)
Charles Bryant, Aculab PLC (U.K.)
Page (NA) Paper number 91
Abstract:
We present a first version of a filter dictionary for use in a computer-telephony text-to-speech synthesis system. The aim of the filter dictionary was to provide a lexicon that was compact, fast and had broader coverage than the standard dictionary used to create it. Correct phonemic transcriptions and lexical stress assignment were both required for a transcription to be deemed accurate. The approach taken here guarantees 100% accurate coverage of the original dictionary, but also gives 93% accurate transcription of the expected coverage of novel words. Lexical stress and the phonemic transcription were retrieved in one pass, resulting in an extremely fast system. We also allowed user-definition to retain accuracy for non-standard transcriptions. This algorithm was developed for British English, but could be applied to other languages.