Incorporating Typographic, Logical and Layout Knowledge of Documents into Text-to-Speech
Γεώργιος Κουρουπέτρογλου · IOS Press eBooks · 2013
Although Text-to-Speech (TtS) is considered a mature technology capable to produce synthetic speech of very high quality, current TtS systems do not include effective acoustic provision of the semantics and the cognitive aspects of the visual (such as the typographic cues) and non-visual (such as the logical structure) knowledge embedded in the rich text documents. In this paper, after the introduction of an appropriate document architecture, we analyze the semantics of the document signals. Then, by following a Design-for-All methodology, we present the Document-to-Audio approach we have developed for the automatics rendering document signals from the typographic, logical and the layout layers to the auditory modality.