Researchers at the University of California, San Diego have developed Dance Dance Convolution, a system that takes as an input the raw audio files of pop songs and produces dance routines as an output--essentially a machine that can choreograph music.
The system initially decides when to place steps, which involves identifying a set of timestamps in a song at which to place steps. Once the timestamps for each step have been identified, the system selects a step to take at each instant.
For this study, the researchers focused on two smaller datasets consisting of recordings plus dance charts, one of which contains 90 songs choreographed by a single author, while the other contains 133 songs each with a single dance chart.
The researchers found the experiments establish the feasibility of using machine learning to automatically generate high-quality dance steps.
From Technology Review
View Full Article
Abstracts Copyright © 2017 Information Inc., Bethesda, Maryland, USA
No entries found