Data-Driven Automatic Choreography
Rongfeng Li · 2024
Advancements in computer vision are revolutionizing the way dance performances are created, and this chapter discusses the impact of technology on the art of choreography. The chapter provides a foundation for subsequent action generation, action filtering, feature matching, and dance synthesis. It introduces methods for analyzing the correlation between music and movement, and used algorithms for selecting and extracting music features. The chapter also includes examples of virtual characters in computer games and commercials. The chapter is divided into several sections, including an introduction to the topic, a discussion of related work, and an overview of the proposed approach. The authors describe the data-driven approach to choreography, which involves analyzing large datasets of dance movements and music to identify patterns and correlations. They also discuss the use of machine learning algorithms to generate new dance movements based on these patterns. The chapter includes several case studies that demonstrate the effectiveness of the proposed approach. These case studies involve the creation of virtual characters that can dance to music in real-time. The authors describe the process of creating these characters, including the selection of music, the extraction of music features, and the generation of dance movements. Overall, this chapter provides a comprehensive overview of the field of data-driven automatic choreography. It highlights the potential of technology to revolutionize the way dance performances are created and provides a foundation for future research in this area. The chapter is a valuable resource for anyone interested in the intersection of technology and art and the potential of data-driven approaches to automate creative processes.