Autonomous Robotic Choreography Creation via Semi-interactive Evolutionary Computation

Hua Peng, Huosheng Hu, Fei Chao, Changle Zhou, Jing Li

Abstract: The creation of robotic choreography is a new hotspot of the artificial intelligence research (Peng in IEEE Trans Hum Mach Syst 45:281–293, 2015), and its final goal is for a robot to create good dance works autonomously. This is a challenging task due to the absence of the objective evaluation criteria. This paper proposes a semi-interactive evolutionary computation method for a humanoid robot to create robotic choreography autonomously. It emphasizes that the humanoid robot should develop its good dance poses and movements autonomously by means of machine learning. The proposed mechanism has been implemented on simulated NAO robots in simulation to verify its feasibility and performance.