Part 2 : Change the music and dance
There are robots which attempt to dance, there are robots which attempt to speak, and there are robots which attempt to respond to voice commands. Far fewer, however, attempt all three at the same time. An exception is the ‘bot created by a joint Japanese / Portuguese university study-group (in association with Honda) led by João Lobato Oliveira of the Artificial Intelligence and Computer Science Laboratory (LIACC) – FEUP, Porto, Portugal.
One of the inherent problems with such a project is that since the robot is producing (pre-recorded) music and replies in a spoken voice, it might sometimes have problems in hearing the voice commands of a nearby human. Thus it was equipped with what the team call an Ego Noise Suppression (ENS) module.
“The experimental tests revealed accurate and robust beat tracking and speech recognition, and convincing dance beat-synchrony.”
- the relative successes of which can be appreciated in the video above. The team’s paper can be read in full here : An active audition framework for auditory-driven HRI: Application to interactive robot dancing
COMING SOON : RO-MAN’s lighter-than-air ‘bot