Online Motion Style Transfer for Interactive Character Control

TL;DR

This paper introduces an end-to-end neural network for real-time online motion style transfer that enables interactive character control in gaming, eliminating handcrafted features and supporting diverse styles.

Contribution

The proposed neural network allows real-time, interactive motion style transfer without handcrafted features, suitable for direct deployment in game systems.

Findings

High accuracy in style transfer

Flexible to different styles and controls

Produces diverse motion outputs

Abstract

Motion style transfer is highly desired for motion generation systems for gaming. Compared to its offline counterpart, the research on online motion style transfer under interactive control is limited. In this work, we propose an end-to-end neural network that can generate motions with different styles and transfer motion styles in real-time under user control. Our approach eliminates the use of handcrafted phase features, and could be easily trained and directly deployed in game systems. In the experiment part, we evaluate our approach from three aspects that are essential for industrial game design: accuracy, flexibility, and variety, and our model performs a satisfying result.