Schlüsselwörter
(Englisch)
|
human-machine communication; information management; interactivity; multimodality; service robots
|
Forschungsprogramme
(Englisch)
|
COST-Action 276 - Information and knowledge management for integrated media communication systems
|
Kurzbeschreibung
(Englisch)
|
The project considers new design methodologies in order to develop intelligent voice-enabled interfaces for human-machine communication, in particular for fully autonomous service robots, within a dynamic adverse environment. Interaction with humans in a natural way can resolve some of the ambiguities in one modality by means of other modalities. This basic research project, with a lot of potential applications, attempts to integrate smart signal acquisition, automatic speech recognition and understanding, speaker verification, speech synthesis and dialogue systems into realistic, information and interactivity management systems for multimodal communication.
|
Weitere Hinweise und Angaben
(Englisch)
|
Full name of research-institution/enterprise: EPF Lausanne STI-IEL-LIDIAP Institut de génie électrique et électronique
|
Partner und Internationale Organisationen
(Englisch)
|
AT, BG, CH, CS, CY, CZ, ES, FR, GR, HR, HU, IE, IT, MK, NO, RO, SI, SK, TR
|
Abstract
(Englisch)
|
The project considers new design methodologies in order to develop voice-enabled interfaces for human-machine communication, in particular for fully autonomous service robots, within a dynamic adverse environment. Interaction with humans in a natural way can resolve some of the ambiguities in one modality by means of other modalities. This basic research project, with a lot of potential applications, attempts to integrate smart signal acquisition, automatic speech recognition and understanding, speech synthesis and dialogue systems into realistic, information and interactivity management system for multimodal communication. The innovative focus of the project is the application of probabilistic (Bayesian) inference and learning to the problem of bahavioural control in autonomous robots. A new technique for modelling unreliable data and a new fusion method for combining modalities, developed in this project, is based on Bayesian networks.
|
Datenbankreferenzen
(Englisch)
|
Swiss Database: COST-DB of the State Secretariat for Education and Research Hallwylstrasse 4 CH-3003 Berne, Switzerland Tel. +41 31 322 74 82 Swiss Project-Number: C03.0034
|
