Abstract
We describe a multimedia installation that provides users with the experience to sing like a tenor from the early 20th century. The user defines vowels with her mouth but does not produce sound. The mouth shape is recognized and tracked by a depth-sensing camera and synthesized using a dedicated sound analysis using formants. Arm gestures are recognized and used to determine pitch and volume of an artificially generated voice. This synthesized voice is additionally modified by acoustic filters to sound like a singing voice from an old gramophone. The installation allows to scan the user's face and to create an individual 3D model of a tenor character that is used to visualize the user performance. © Springer International Publishing 2013.
Cite
CITATION STYLE
Feitsch, J., Strobel, M., & Geiger, C. (2013). Singing like a tenor without a real voice. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 8253 LNCS, pp. 258–269). https://doi.org/10.1007/978-3-319-03161-3_18
Register to see more suggestions
Mendeley helps you to discover research relevant for your work.
