English

Transflower: probabilistic autoregressive dance generation with multimodal attention

Sound 2022-06-14 v3 Graphics Machine Learning Audio and Speech Processing

Abstract

Dance requires skillful composition of complex movements that follow rhythmic, tonal and timbral features of music. Formally, generating dance conditioned on a piece of music can be expressed as a problem of modelling a high-dimensional continuous motion signal, conditioned on an audio signal. In this work we make two contributions to tackle this problem. First, we present a novel probabilistic autoregressive architecture that models the distribution over future poses with a normalizing flow conditioned on previous poses as well as music context, using a multimodal transformer encoder. Second, we introduce the currently largest 3D dance-motion dataset, obtained with a variety of motion-capture technologies, and including both professional and casual dancers. Using this dataset, we compare our new model against two baselines, via objective metrics and a user study, and show that both the ability to model a probability distribution, as well as being able to attend over a large motion and music context are necessary to produce interesting, diverse, and realistic dance that matches the music.

Keywords

Cite

@article{arxiv.2106.13871,
  title  = {Transflower: probabilistic autoregressive dance generation with multimodal attention},
  author = {Guillermo Valle-Pérez and Gustav Eje Henter and Jonas Beskow and André Holzapfel and Pierre-Yves Oudeyer and Simon Alexanderson},
  journal= {arXiv preprint arXiv:2106.13871},
  year   = {2022}
}

Comments

Article presented at SIGGRAPH Asia 2021, and published in ACM Transactions on Graphics

R2 v1 2026-06-24T03:37:02.048Z