Complex Emotions - The Simultaneous Simulation of Emotion- Related States In Synthesized Speech


We describe an approach to simulate first and secondary emotional expression in synthesized speech simultaneously by targeting different parameter categories. The approach is based on the open-source system “Emofilt” which utilizes the diphone-synthesizer “Mbrola”. The evaluation of the approach by a perception experiment showed that the pure emotions were all recognized above chance. Whereas the results are promising, the ultimate aim to validly synthesize two emotions simultaneously was not fully reached. Apparently, some emotions dominate the perception (fear), and the salience or quality of synthesis does not seem to be equally distributed over the two feature bundles.

Year: 2017
In session: Poster
Pages: 67 to 74