May the Dance be with You: Dance Generation Framework for Non-Humanoids

TL;DR

This paper introduces a novel framework enabling non-humanoid agents to learn dance movements from human videos by aligning optical flow with music through reinforcement learning and contrastive feature encoding.

Contribution

It presents the first framework for non-humanoid dance generation from human videos, utilizing a reward model based on contrastive learning of optical flow and music features.

Findings

Generated dances align well with music beats

The framework outperforms baseline methods in human preference

Optical flow effectively captures visual rhythm for dance synthesis

Abstract

We hypothesize dance as a motion that forms a visual rhythm from music, where the visual rhythm can be perceived from an optical flow. If an agent can recognize the relationship between visual rhythm and music, it will be able to dance by generating a motion to create a visual rhythm that matches the music. Based on this, we propose a framework for any kind of non-humanoid agents to learn how to dance from human videos. Our framework works in two processes: (1) training a reward model which perceives the relationship between optical flow (visual rhythm) and music from human dance videos, (2) training the non-humanoid dancer based on that reward model, and reinforcement learning. Our reward model consists of two feature encoders for optical flow and music. They are trained based on contrastive learning which makes the higher similarity between concurrent optical flow and music features. With this reward model, the agent learns dancing by getting a higher reward when its action creates an optical flow whose feature has a higher similarity with the given music feature. Experiment results show that generated dance motion can align with the music beat properly, and user study result indicates that our framework is more preferred by humans compared to the baselines. To the best of our knowledge, our work of non-humanoid agents which learn dance from human videos is unprecedented. An example video can be found at https://youtu.be/dOUPvo-O3QY.