Segmentation, Labelling and Speech Corpora 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
Acoustic Indicators Of Topic Segmentation
Authors:
Julia Hirschberg, AT&T Labs / Research (USA)
Christine H. Nakatani, AT&T Labs / Research (USA)
Page (NA) Paper number 976
Abstract:
The segmentation of text and speech into topics and subtopics is an important step in document interpretation. For text, formatting information, such as headings and paragraphing, is available to aid in this endeavor, although this information is by no means sufficient. For speech, the task is even more difficult. We present results of the application of machine learning techniques to the automatic identification of intonational phrases beginning and ending 'topics' determined independently by annotators for two corpora | the Boston Directions Corpus and the Broadcast News (HUB-4) DARPA/NIST database.
SL980976.PDF (From Author)
SL980976.PDF (Rasterized)
TOP
IViE - A Comparative Transcription system for Intonational Variation in English
Authors:
Esther Grabe, University of Cambridge (U.K.)
Francis Nolan, University of Cambridge (U.K.)
Kimberley J. Farrar, University of Cambridge (U.K.)
Page (NA) Paper number 99
Abstract:
In this paper, we offer an alternative to ToBI, the current de facto standard for machine-readable labelling of English prosody. We have three reasons for arguing that an alternative is needed. Firstly, the ToBI tone inventory is not maximally constrained; it appears to be difficult for transcribers to reach high inter-transcriber agreement scores for tone labels. Secondly, the growing demand for prosodically labelled data from non-standard varieties of English suggests a need for a transparent comparative transcription system. ToBI was not designed for this purpose. Thirdly, the low inter-transcriber agreement scores for ToBI suggest that the system is not as easy to apply as it may at first appear. In the present paper, we describe an alternative: the IViE system (Intonational Variation in English). We describe the structure of IViE and discuss its application with examples.
SL980099.PDF (From Author)
SL980099.PDF (Rasterized)
TOP
Automatic Segmental and Prosodic Labeling of Mandarin Speech Database
Authors:
Fu-Chiang Chou, Dept. of Electrical Engineering, National Taiwan University (Taiwan)
Chiu-Yu Tseng, Institute of Linguistics, Preparatory Office, Academia Sinica (Taiwan)
Lin-Shan Lee, Dept. of Electrical Engineering, National Taiwan University (Taiwan)
Page (NA) Paper number 266
Abstract:
In this paper we describe the techniques and methodology developed for automatic labeling of segmental and prosodic information for the Mandarin speech database. There are two major procedures. First, the text is converted into the phonetic network of possible pronunciations, and this network is aligned with the speech data by recognition processes. Secondly, many acoustic prosodic features are derived and the break indices are labeled with these features by decision trees. For the segmental labeling, 96.5% of automatically determined segment boundaries are accurate within a range of 20 ms. For the prosodic labeling, 84.9% of the automatic labeled break indices are the same with the manual labeled one.
SL980266.PDF (From Author)
SL980266.PDF (Rasterized)
TOP
Automatic Labelling of German Prosody
Authors:
Stefan Rapp, Sony International (Europe) GmbH (Germany)
Page (NA) Paper number 907
Abstract:
We present research on an automatic labelling system that is able to produce a phonological tonal labelling according to the ToBI like intonation model for German developed by Fery. The current system was trained on about 1 hour of expert prosodically labelled speech from a single male radio news announcer. We present experiments for finding a suitable feature set drawn from features that describe the prosodic correlates fundamental frequency, duration and intensity as well as some lexical and syntactic features. With the best feature set, we achieve a recognition rate of 78.7% for speaker dependent recognition of ToBI labels (simultaneously predicting prominence and phrasing) and 86.9% for the simpler accented/not accented decision. Although the system's accuracy is well below that of human transcribers, it is a useful tool actively used in our laboratory due to it's ability to process large amounts of speech data at low costs.