Beyond Imitation: Generative and Variational Choreography via Machine Learning
This work addresses the challenge of automated choreography creation for artists and performers, representing an incremental advancement in applying existing ML methods to a new domain.
The researchers tackled the problem of generating novel choreography sequences and creating tunable variations on existing ones using machine learning, resulting in the development of configurable tools based on recurrent neural networks and autoencoders trained on movement data.
Our team of dance artists, physicists, and machine learning researchers has collectively developed several original, configurable machine-learning tools to generate novel sequences of choreography as well as tunable variations on input choreographic sequences. We use recurrent neural network and autoencoder architectures from a training dataset of movements captured as 53 three-dimensional points at each timestep. Sample animations of generated sequences and an interactive version of our model can be found at http: //www.beyondimitation.com.