AVALIAÇÃO DE RECONHECEDORES DE VOZ PARA PORTUGUÊS BRASILEIRO NA PLATAFORMA ANDROID

Abstract

The growing use of mobile devices has providing an increase in digital inclusion, however people with disabilities are not fully included in this scenario because there are few applications that promote accessibility. This problem can be solved by the use of different interface types. In this scenario, the technology of Automatic Speech Recognition (ASR) can be used in natural interfaces of interaction human-computer. In this context, the aim of this work is to show the performance comparison between the speech recognizers for Brazilian Portuguese (BP) Google Android and Nuance Dragon on converters applications from Speech to Text (STT) for mobile devices based on Android Operating System. Performance tests were made with users divided into groups by age, where users tested in real environments two applications, each implementing voice recognizers mentioned. The goal of STT applications is to provide accessibility for people with physical disabilities to computational systems. Index Terms - Accessibility, Mobile Device, Speech Recognition, Tests.