Affective Speech for Social Communication: Implementation Challenges in Text-to-Speech for Short Messages

The flexibility to choose from different modal content presentation will be an important feature in future ubiquitous application. Currently, short messages (e.g. SMS/MMS) are only available in visual form. However, in certain situations, users may like to have these messages presented in audio form. We explored the alternative of presenting short messages in affective synthetic text-to-speech form special for social communications between teens. Evaluation of this alternative presentation reveals that, for emotion recognition, it was easier to interpret emotion messages generated from affective synthetic speech. Although there is no actual difference in the way people think they were able to derive emotions from both types of messages. For teens, affective synthetic speech is sometimes fun to usei