Speech processing for interactive pronunciation learning : from prototyping environments to scalable «web» services

Details

Supervisors

Macq, Benoît

Faculty

Ecole polytechnique de Louvain

Degree label

Master [120] en sciences informatiques

Abstract

GRAAL is a mobile application that offers a self-learning or classroom learning environment for English pronunciation acquisition. It uses speech analysis and speech processing through immersive stories and gamified activities for interactive pronunciation learning. Intended for global deployment and availability over the network, the GRAAL application needs a solid back-end server capable of accepting high loads of analysis requests. Consequently, the purpose of this thesis is to develop a prototype of such an analysis server capable of meeting the computing power and market's demand for the GRAAL application. The proposed solution is in the form of interconnected architecture of micro-services deployed at a cloud service provider.