Synthesis of Compositional Animations from Textual Descriptions

TL;DR

This paper introduces a hierarchical two-stream model that generates realistic 3D human motion sequences from complex natural language descriptions, advancing the field of text-based motion synthesis.

Contribution

It presents a novel hierarchical model that effectively maps complex sentences to detailed 3D pose sequences, handling compositional actions and improving synthesis quality.

Findings

Achieved a 50% improvement over previous methods in objective evaluations.

Generated motions closely matched ground-truth data in user studies.

Successfully handled both simple and complex, multi-action sentences.

Abstract

"How can we animate 3D-characters from a movie script or move robots by simply telling them what we would like them to do?" "How unstructured and complex can we make a sentence and still generate plausible movements from it?" These are questions that need to be answered in the long-run, as the field is still in its infancy. Inspired by these problems, we present a new technique for generating compositional actions, which handles complex input sentences. Our output is a 3D pose sequence depicting the actions in the input sentence. We propose a hierarchical two-stream sequential model to explore a finer joint-level mapping between natural language sentences and 3D pose sequences corresponding to the given motion. We learn two manifold representations of the motion -- one each for the upper body and the lower body movements. Our model can generate plausible pose sequences for short sentences describing single actions as well as long compositional sentences describing multiple sequential and superimposed actions. We evaluate our proposed model on the publicly available KIT Motion-Language Dataset containing 3D pose data with human-annotated sentences. Experimental results show that our model advances the state-of-the-art on text-based motion synthesis in objective evaluations by a margin of 50%. Qualitative evaluations based on a user study indicate that our synthesized motions are perceived to be the closest to the ground-truth motion captures for both short and compositional sentences.