About This Project
These dance clips were generated using EDGE (Editable Dance GEneration), a generative AI model developed by Stanford University researchers that choreographs realistic human dance animations synchronized to any piece of music. (Stanford HAI, April 2023)
EDGE analyzes a song's rhythmic and emotional content and produces physically plausible choreography — moves a real dancer could actually perform — without any manual frame-by-frame animation. Its key advantage is editability: a choreographer can fix specific limb positions and EDGE auto-completes the rest of the body in a way that stays true to the music, making iterative creative refinement fast and intuitive.
For this project, EDGE was applied to kid-friendly music — upbeat, playful tracks — so the generated choreography is energetic and approachable for young learners. The goal is to make dance teaching more engaging: kids can watch a character move to music they already love, making it easier and more fun to follow along and learn new moves.
After generation, the raw clips were post-processed in 3D software to refine lighting, color, and character appearance, giving the animations a polished, visually appealing look that is more attractive and exciting for kids.