SpeakerID - Toolkit para Processamento e Modelagem de Características de Alto Nível para Reconhecimento Automático de Locutor

Continuous Interaction with TDK: Improving the User Experience in Terralib

Cristian Keil de AbreuAndré Gustavo Adami

Marcelo MetelloMário de Sá VeraMelissa LemosLeone Pereira MasieroMarcelo Tilio Monteiro de Carvalho

This paper describes a toolkit for feature extraction and modeling of high-level features for speaker recognition systems. In addition, the toolkit provides tools to evaluate features and models according to the NIST evaluation paradigm (commonly used as reference for evaluating speaker recognition systems). The toolkit was implemented in Perl and C languages and uses several open-source software for process scheduling and feature extraction. Some results of the toolkit on the 2001 NIST competition are presented in this paper. Historically, visual display has always played a very important role in GIS applications. However, visual exploration tools do not scale well when applied to huge spatial data sets. More recently, faster processing hardware and more sophisticated computer graphics has been used to improve user experience for geospatial data visualization. Some applications have recently introduced the concept that the user should be able to navigate through the data in a smooth and continuous way without being blocked waiting for data to be loaded, even if this means seeing incomplete data for some time. Google Earth and NASA World Wind are recent examples of this concept. This paper describes an architecture incorporated in the TDK (TerraLib Development Kit) open source API to provide support for applications that want this kind of user interaction with Terralib geographical data.

Caso o link acima esteja inválido, faça uma busca pelo texto completo na Web: Buscar na Web

Biblioteca Digital Brasileira de Computação - Contato:
     Mantida por: