Spoken Language Models and Dialog 3
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
On Different Functions of Repetitive Utterances
Authors:
Marc Swerts, IPO, Center for Research on User-System Interaction (The Netherlands)
Hanae Koiso, ATR, Media Integration and Communications Research Laboratories (Japan)
Atsushi Shimojima, ATR, Media Integration and Communications Research Laboratories (Japan)
Yasuhiro Katagiri, ATR, Media Integration and Communications Research Laboratories (Japan)
Page (NA) Paper number 269
Abstract:
The study reported in this paper focuses on different functions of echoing in Japanese dialogues. Echoing is defined as a speaker's lexical repeat of (parts of) an utterance spoken by a conversation partner in a previous turn. The phenomenon was investigated in three task-oriented, informal dialogues. Repeats in this corpus were labeled in terms of a 5-point scale which expressed the level to which a speaker had integrated the other person's utterance into his/her own body of knowledge. Kappa statistics showed that the labels could reliably be reproduced by three independent subjects. The investigation brought to light that the level of integration is reflected in a number of lexical and prosodic correlates. These features are discussed regarding their information potential, i.e., their accuracy and comprehensiveness as signals.
SL980269.PDF (From Author)
SL980269.PDF (Rasterized)
TOP
Prosody-Based Detection of the Context of Backchannel Responses
Authors:
Hiroaki Noguchi, Nara Institute of Science and Technology (Japan)
Yasuharu Den, Nara Institute of Science and Technology (Japan)
Page (NA) Paper number 662
Abstract:
Current spoken dialogue systems lack positive feedback such as backchannels, which are common in human-human conversations. To develop more natural human-computer interfaces, the investigation of backchannel-responses are indispensable. In this paper, we propose a method for detecting the precise timing for backchannel responses in Japanese and aim at incorporating such method in future spoken dialogue systems. The proposed method is based on machine learning technique with a variety of prosodic features. It is shown to be effective in automatically deriving rules for detecting the contexts of backchannels. The performance of our method is considerably better than previous methods.
SL980662.PDF (From Author)
SL980662.PDF (Rasterized)
TOP
Robust Interpretation for Spoken Dialogue Systems
Authors:
Lena Strömbäck, Department of Computer and Information Science, Linköping University (Sweden)
Arne Jönsson, Department of Computer and Information Science, Linköping University (Sweden)
Page (NA) Paper number 478
Abstract:
Spoken dialogue systems must allow for robust and efficient interpretation of user utterances. This can be achieved by using shallow and partial interpretation. Partial interpretation is feasible together with a dialogue manager which provides information to guide the analysis. In this paper we present results on developing interfaces for information retrieval applications utilizing partial and information directed interpretation with unification-based formalisms, traditionally used for deep and complete analysis. The major advantage with our approach is that the time to develop the interpretation modules is reduced. Furthermore, the system will be fairly robust as large parts of the knowledge bases containing knowledge on ways in which a user can express a domain concept can be generated automatically or semi-automatically.
SL980478.PDF (From Author)
SL980478.PDF (Rasterized)
TOP
System-User Interaction and Response Strategy in Spoken Dialogue System
Authors:
Yohei Okato, University of Tsukuba (Japan)
Keiji Kato, University of Tsukuba (Japan)
Mikio Yamamoto, University of Tsukuba (Japan)
Shuichi Itahashi, University of Tsukuba (Japan)
Page (NA) Paper number 683
Abstract:
There are a number of restrictions in human-machine interactions, which continue to warrant a better control of response utterances in spoken dialogue systems. In this paper, we investigated how system-response strategies influence users, focusing on: interjectory responses to users' utterances; and on verbose responses which vary between brief and detailed expression used for confirmation. The dialogue task selected consists of telephone shopping, and human-machine dialogues using Wizard-of-OZ method were collected for analysis. The results show that back-channel feedbacks and brief confirmations by our spoken dialogue system, prompt more utterances from and give more satisfaction to users. Second, users tend to give back-channel feedbacks whenever the system gives them feedbacks. It follows that users are able to gauge the system's ability to handle interjections and to predict its behaviour.
SL980683.PDF (From Author)
SL980683.PDF (Rasterized)
TOP
Organizing Self-Motivated Dialogue with Autonomous Creatures
Authors:
Noriko Suzuki, ATR MI&C (Japan)
Kazuo Ishii, ATR MI&C (Japan)
Michio Okada, ATR MI&C (Japan)
Page (NA) Paper number 1101
Abstract:
This paper discusses the effectiveness of human-computer interaction in our prototype system "Talking Eye", which is based on social behaviors in self-motivated dialogue. Talking Eye system consists of autonomous creatures having the emergent computation architecture with the advantage of self- motivated dialogue. We performed an experiment using subject's impression to illustrate the effectiveness of self- motivated dialogue with the Talking Eye system. The result was obtained that they could find personal nature during interaction even if they could not accomplish their conversational purpose. The main goal of this work is to build a mechanism for autonomous creatures that promotes a more consensual feeling through its interaction with humans.
SL981101.PDF (From Author)
SL981101.PDF (Rasterized)
TOP
Fly with the EAGLES: Evaluation of the "ACCeSS" Spoken Language Dialogue System
Authors:
Gerhard Hanrieder, Daimler-Benz Aerospace AG (Germany)
Paul Heisterkamp, Daimler-Benz AG (Germany)
Thomas Brey, University of Regensburg (Germany)
Page (NA) Paper number 274
Abstract:
This paper reports the experiences we had in evaluating the ACCeSS system using the EAGLES evaluation metrics both at the input/output (black box evaluation) and component levels (glass box evaluation). Our primary motivation for using the EAGLES metrics was to apply a set of metrics that allow objective measurement of system performance. These measures are used both for diagnostic evaluation of the current version and for progress evaluation of successive system versions. The EAGLES metrics provided very valuable guidelines for our evaluation, but not all the metrics were unproblematic. Especially for diagnostics during development, the addition of subtask success rate proved very helpful.