LOLOL: Laugh Out Loud On Laptop

"Significant progress in the domains of speech- and singing-synthesis hasenhanced communicative potential of machines. To make computers more vocallyexpressive, however, we need a deeper understanding of how nonlinguistic socialsignals are patterned and perceived. In this paper, we focus on laughterexpressions: how a phrase of vocalized notes that we call ""laughter"" may bemodeled and performed to implicate nuanced meaning imbued in the acousticsignal. In designing our model, we emphasize (1) using high-level descriptorsas control parameters, (2) enabling real-time performable laughter, and (3)prioritizing expressiveness over realism. We present an interactive systemimplemented in ChucK that allows users to systematically play with the musicalingredients of laughter. A crowdsourced study on the perception of synthesizedlaughter showed that our model is capable of generating a range of laughtertypes, suggesting an exciting potential for expressive laughter synthesis."