Recent Advances in the Application of Deep Learning to Choreography

Choreography is the arrangement and design of dance movements according to music. The application of computer technology to dance generation improves the efficiency and quality of choreography. This paper categorizes the methods used in dance generation over the last two decades. And the paper summarizes the latest research issues in this field, including creating new movements, matching the movement to the rhythm, conducting choreograph conservation, and finding the semantic connection between dance and music. The researchers' models and results for these issues are detailed in the paper. Finally, the future research direction is discussed.