Using Audio and Haptic Feedback to Improve Pitched Percussive Instrument Performance in Humanoids

Batula, Alyssa and Colacot, Manu and Grunberg, David and Kim, Youngmoo

Proceedings of the International Conference on New Interfaces for Musical Expression

We present a system which allows an adult-sized humanoid to determine whetheror not it is correctly playing a pitched percussive instrument to produce adesired sound. As hu, man musicians utilize sensory feedback to determine ifthey are successfully using their instruments to generate certain pitches,robot performers should be capable of the same feat. We present a noteclassification algorithm that uses auditory and haptic feedback to decide if anote was well- or poorly-struck. This system is demonstrated using Hubo, anadult-sized humanoid, which has been enabled to actu, ate pitched pipes usingmallets. We show that, with this system, Hubo is able to determine whether ornot a note was played correctly.