Harnessing generative neural networks to fuse traditional Tujia Baishou dance with contemporary choreography: Enhancing creativity and aesthetic experience in dance students

Abstract

The objective of this study was to explore the potential of using the generative neural network Dance2Dance to integrate elements of the traditional Tujia Baishou dance with modern choreography, specifically examining the impact of such technologies on the creative and improvisational skills of dance students. The findings indicate that the use of dance sequences generated by the Dance2Dance neural network increased the originality of dance performances by the experimental group students, with scores rising from 3.45 to 4.48 out of a maximum of 5.00 (W = 182.50, p = 0.000, according to the Wilcoxon test), whereas the control group's scores remained unchanged. The Aesthetic Experiences Scale, which measures unique and intense experiences with art, revealed a higher level of positive aesthetic perception in the experimental group (3.36) compared to the control group (2.11). Interviews assessing the compositions created using the generative neural network for their balance between contemporary and Baishou dance elements, mutual compatibility, and choreographic skill indicated that the modernized choreography generally appeared more professional than the original Tujia Baishou dance. However, some shortcomings were identified and analyzed. The results obtained demonstrate the potential of using generative neural networks to create new dance sequences that combine features of various stylistic directions, as well as to develop creative and improvisational skills. These findings may be valuable for choreographers and dance instructors considering the integration of AI technologies to enhance the choreography creation process.

Keywords: AI; Dance training; Digital choreographies; Generative neural networks; Technologies for dancers; Traditional Chinese dances.