Segmentation, Labelling and Speech Corpora 4
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 Multilingual Prosodic Database
Authors:
Estelle Campione, Université de Provence (France)
Jean Véronis, Université de Provence (France)
Page (NA) Paper number 844
Abstract:
We present a prosodic corpus in five languages (French, English, Italian, German and Spanish) comprising 4 hours and 20 minutes of speech and involving 50 different speakers (5 male and 5 female per language). The recordings on which the corpus is based are extracted from the EUROM 1 speech database and consists of passages of about five sentences. The corpus was stylized automatically by an algorithm which factors out microprosodic effects and represents the intonation contour of utterances by a series of target points. Once interpolated by a smooth curve (spline), these points produce a contour undistinguishable from the original when re-synthesized, apart from a few detection errors. A symbolic coding of the 50000 pitch movements of the corpus is also provided, along with the time-alignment of orthographic transcription to signal at word-level. The entire corpus was verified and manually corrected by experts for each language. It will be made available at production cost for research through the European Language Resource Association (ELRA).
SL980844.PDF (From Author)
SL980844.PDF (Rasterized)
TOP
The CSLU Speaker Recognition Corpus
Authors:
Ronald A. Cole, CSLU (USA)
Mike Noel, CSLU (USA)
Victoria Noel, CSLU (USA)
Page (NA) Paper number 856
Abstract:
This paper describes the CSLU Speaker Recognition Corpus data collection. The corpus was motivated by a need for speech data from many speakers, under different environmental conditions, with each speaker providing data over a significant period of time. The corpus was designed to provide sufficient data to study phonetic variability within and across sessions, and to design and evaluate systems for both vocabulary independent and vocabulary specific recognition and verification tasks. The protocol includes fixed vocabulary phrases, digit strings, personal utterances (e.g., eye color), and fluent speech. The resulting Speaker Recognition Corpus is a collection of telephone speech recordings from over 500 participants collected over a two-year period. We describe the data collection procedure, the protocol, the transcription methods and the current status of the Speaker Recognition Corpus.
SL980856.PDF (From Author)
SL980856.PDF (Rasterized)
TOP
How Effective Is Unsupervised Data Collection For Children's Speech Recognition?
Authors:
Gregory Aist, Carnegie Mellon University (USA)
Peggy Chan, Carnegie Mellon University (USA)
Xuedong Huang, Microsoft Research (USA)
Li Jiang, Microsoft Research (USA)
Rebecca Kennedy, Carnegie Mellon University (USA)
DeWitt Latimer, Carnegie Mellon University (USA)
Jack Mostow, Carnegie Mellon University (USA)
Calvin Yeung, Carnegie Mellon University (USA)
Page (NA) Paper number 929
Abstract:
Children present a unique challenge to automatic speech recognition. Today's state-of-the-art speech recognition systems still have problems handling children's speech because acoustic models are trained on data collected from adult speech. In this paper we describe an inexpensive way to mend this problem. We collected children's speech when they interact with an automated reading tutor. These data are subsequently transcribed by a speech recognition system and automatically filtered. We studied how to use these automatically collected data to improve children's speech recognition system's performance. Experiments indicate that automatically collected data can reduce the error rate significantly on children's speech.
SL980929.PDF (From Author)
SL980929.PDF (Rasterized)
TOP
An Algorithm for Automatic Generation of Mandarin Phonetic Balanced Corpus
Authors:
Jyh-Shing Shyuu, Department of Computer Science and Information Engineering (Taiwan)
Wang Jhing-Fa, Department of Computer Science and Information Engineering (Taiwan)
Page (NA) Paper number 960
Abstract:
This paper proposed an algorithm for automatic generation of Mandarin phonetic balanced corpus. The design of phonetic balanced corpus is particularly important for the collection of continuous speech database to reduce the co-articulate effects in continuous speech recognition(CSR). Traditionally, balanced corpus is generated manually or semi- automatically. Our proposed algorithm tries to find a minimum number of sentences from a large text corpus set and ensures that 408 Mandarin base syllables and 38*22 co-articulations between vowels and consonants are distributed in the extracted sentences.
SL980960.PDF (From Author)
SL980960.PDF (Rasterized)
TOP
Towards a Formal Framework for Linguistic Annotations
Authors:
Steven Bird, LDC, University of Pennsylvania (USA)
Mark Liberman, LDC, University of Pennsylvania (USA)
Page (NA) Paper number 774
Abstract:
`Linguistic annotation' is a term covering any transcription, translation or annotation of textual data or recorded linguistic signals. While there are several ongoing efforts to provide formats and tools for such annotations and to publish annotated linguistic databases, the lack of widely accepted standards is becoming a critical problem. Proposed standards, to the extent they exist, have focussed on file formats. This paper focuses instead on the logical structure of linguistic annotations. We survey a wide variety of annotation formats and demonstrate a common conceptual core. This provides the foundation for an algebraic framework which encompasses the representation, archiving and query of linguistic annotations, while remaining consistent with many alternative file formats.
SL980774.PDF (From Author)
SL980774.PDF (Rasterized)
TOP
Forming Generic Models Of Speech For Uniform Database Access
Authors:
Toomas Altosaar, Helsinki University of Technology (Finland)
Martti Vainio, University of Helsinki (Finland)
Page (NA) Paper number 887
Abstract:
This paper presents a formalism that models speech from different databases generically. For each utterance in a speech database a communication framework is first constructed which is composed of a set of communication planes, such as acoustic, orthographic, linguistic, and phonetic. Each plane in turn is made up of a set of levels to represent the plane's structural hierarchy, e.g., for the linguistic plane, levels such as sentence, word, syllable, and phoneme may exist. Information from speech databases is parsed and compiled into such objects and exhibit both individual and class inherited behaviour. Once placed into the framework these objects can have their relationships to other objects explicitly defined by links on the same level, across different levels, and across different planes. Speech from databases covering different languages and annotation styles can therefore be modelled generically allowing for uniform database access. Searches can be performed on the framework and the results used for further analyses.