Learn More
This paper describes the context-aware mobile tourist application COMPASS that adapts its services to the user's needs based on both the user's interests and his current context. In order to provide context-aware recommendations, a recommender system has been integrated with a context-aware application platform. We describe how this integration has been(More)
Service discovery is a process of locating, or discovering, one or more documents, that describe a particular service. Most of the current service discovery approaches perform syntactic matching, that is, they retrieve services descriptions that contain particular keywords from the user's query. This often leads to poor discovery results, because the(More)
The web has evolved from the HTML-based repositories of text and images towards the current web services and semantic web developments. Web services are believed to help the integration of diverse applications while the semantic web promises to increase the " intelligence " of the web, enabling richer discovery, data integration, navigation and automation(More)
This paper summarizes the main results from the Speaker Veriication (SV) research pursued so far in the CAVE project. Diierent state-of-the art SV algorithms were implemented in a common HMM framework and compared on two databases : YOHO (of-ce environment speech) and SESP (telephone speech). This paper is concerned with the diierent design issues for(More)
The issue of a priori threshold setting in speaker veri-cation is a key problem for eld applications. In the context of the CAVE project, we compared several methods for estimating speaker-independent and speaker-dependent decision thresholds. Relevant parameters are estimated from development data only, i.e. without resorting to additional client data. The(More)
The problem of how to train variance parameters on scarce data is addressed in the context of text-dependent, HMM-based, automatic speaker verification. Three variations of variance flooring is explored as a means to prevent over-fitting. With the best performing one, the floor to a variance vector of a client model is proportional to the corresponding(More)
R ESUM E Cet article pr esente un panorama des activit es de recherche men ees dans le cadre du projet europ een LE-Telematics CAVE, sur la v eriication du locuteur a travers le r eseau t el ephonique. Nous pr esentons tout d'abord les principes des algorithmes sur lesquels repose notre syst eme de v eri-cation. Nous d ecrivons ensuite le protocole et la(More)
In this paper we present a web services-based platform that facilitates and speeds up the development and deployment of context-aware, integrated mobile speech and data applications. The platform is capable of handling different types of context and offers sophisticated personalization mechanisms. To illustrate the usefulness of the platform and to validate(More)
This paper addresses diierences between operational speaker veriication (SV) systems and laboratory experiments in terms of performance and methods for measuring performance. It is concluded that operational SV systems need an indication of the quality of newly enrolled speaker models , to decide whether to re-enrol or request more enrolment material. We(More)