Generating Attribute-Aware Human Motions from Textual Prompt

TL;DR

This paper introduces a novel framework for generating human motions from text that explicitly incorporates human attributes like age and gender, enabling more personalized and accurate motion synthesis.

Contribution

It proposes a new attribute-aware motion generation model based on Structural Causal Models, addressing the gap of attribute influence in text-driven human motion synthesis.

Findings

The model effectively decouples action semantics from human attributes.

A new dataset with attribute annotations is introduced as a benchmark.

Experiments demonstrate the model's ability to generate personalized human motions.

Abstract

Text-driven human motion generation has recently attracted considerable attention, allowing models to generate human motions based on textual descriptions. However, current methods neglect the influence of human attributes-such as age, gender, weight, and height-which are key factors shaping human motion patterns. This work represents a pilot exploration for bridging this gap. We conceptualize each motion as comprising both attribute information and action semantics, where textual descriptions align exclusively with action semantics. To achieve this, a new framework inspired by Structural Causal Models is proposed to decouple action semantics from human attributes, enabling text-to-semantics prediction and attribute-controlled generation. The resulting model is capable of generating attribute-aware motion aligned with the user's text and attribute inputs. For evaluation, we introduce a comprehensive dataset containing attribute annotations for text-motion pairs, setting the first benchmark for attribute-aware motion generation. Extensive experiments validate our model's effectiveness.