MultiAct: Long-Term 3D Human Motion Generation from Multiple Action Labels

TL;DR

MultiAct is a novel framework that generates realistic long-term 3D human motions from multiple action labels, effectively handling transitions and user control, surpassing previous methods' limitations.

Contribution

It introduces the first unified recurrent system for long-term motion generation from multiple action labels, enabling controllable and seamless action sequences.

Findings

Produces realistic long-term motions with smooth transitions.

Effectively controls motion sequences based on multiple user-defined actions.

Outperforms previous action- and motion-conditioned methods.

Abstract

We tackle the problem of generating long-term 3D human motion from multiple action labels. Two main previous approaches, such as action- and motion-conditioned methods, have limitations to solve this problem. The action-conditioned methods generate a sequence of motion from a single action. Hence, it cannot generate long-term motions composed of multiple actions and transitions between actions. Meanwhile, the motion-conditioned methods generate future motions from initial motion. The generated future motions only depend on the past, so they are not controllable by the user's desired actions. We present MultiAct, the first framework to generate long-term 3D human motion from multiple action labels. MultiAct takes account of both action and motion conditions with a unified recurrent generation system. It repetitively takes the previous motion and action label; then, it generates a smooth transition and the motion of the given action. As a result, MultiAct produces realistic long-term motion controlled by the given sequence of multiple action labels. Codes are available here at https://github.com/TaeryungLee/MultiAct_RELEASE.