Spoken notifications in smart environments using Croatian language

Speech technologies have advanced significantly in the last decade, mostly due to rise available computing power combined with novel approaches natural language processing. As a result, speech-enabled systems become popular commercial products, successfully integrated various environments. However, this can be stated for English and few other ?big? languages. From perspective of minority language, such as Croatian, there are many challenges ahead achieve comparable results. In paper, we propose model generation speech synthesis smart environment using Croatian language. The is evaluated on 27 users estimate quality user experience. evaluation goal was determine what perceive more important ? generated or grammatical correctness spoken content. It shown that most perceived grammatically correct texts being highest quality.