Handysinger : Expressive Singing Voice Morphing Using Personified Hand-Puppet Interface
Tomoko Yonezawa, Suzuki, Takahiko, Kenji Mase, Kiyoshi Kogure
Abstract
Open-access reader
Tomoko Yonezawa, Suzuki, Takahiko, Kenji Mase, Kiyoshi Kogure
Abstract
Open-access reader
The HandySinger system is a personified tool developedto naturally express a singing voice controlled by the gestures of a hand puppet. Assuming that a singing voice is akind of musical expression, natural expressions of the singingvoice are important for personification. We adopt a singingvoice morphing algorithm that effectively smoothes out thestrength of expressions delivered with a singing voice. Thesystem's hand puppet consists of a glove with seven bendsensors and two pressure sensors. It sensitively capturesthe user's motion as a personified puppet's gesture. Tosynthesize the different expressional strengths of a singingvoice, the ``normal'' (without expression) voice of a particular singer is used as the base of morphing, and three different expressions, ``dark,'' ``whisper'' and ``wet,'' are used asthe target. This configuration provides musically expressedcontrols that are intuitive to users. In the experiment, weevaluate whether 1) the morphing algorithm interpolatesexpressional strength in a perceptual sense, 2) the handpuppet interface provides gesture data at sufficient resolution, and 3) the gestural mapping of the current systemworks as planned.
OpenAlex reports 13 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
The HandySinger system is a personified tool developedto naturally express a singing voice controlled by the gestures of a hand puppet. Assuming that a singing voice is akind of musical expression, natural expressions of the singingvoice are important for personification. We adopt a singingvoice morphing algorithm that effectively smoothes out thestrength of expressions delivered with a singing voice. Thesystem's hand puppet consists of a glove with seven bendsensors and two pressure sensors. It sensitively capturesthe user's motion as a personified puppet's gesture. Tosynthesize the different expressional strengths of a singingvoice, the ``normal'' (without expression) voice of a particular singer is used as the base of morphing, and three different expressions, ``dark,'' ``whisper'' and ``wet,'' are used asthe target. This configuration provides musically expressedcontrols that are intuitive to users. In the experiment, weevaluate whether 1) the morphing algorithm interpolatesexpressional strength in a perceptual sense, 2) the handpuppet interface provides gesture data at sufficient resolution, and 3) the gestural mapping of the current systemworks as planned.
Key concepts: Singing, Morphing, Computer science, Gesture, Expression (computer science), Interface (matter), Speech recognition, Human voice