Coordinate-Based Dual-Constrained Autoregressive Motion Generation

TL;DR

This paper introduces CDAMD, a novel autoregressive framework for text-to-motion generation that enhances fidelity and semantic consistency by using coordinate inputs and dual constraints, establishing new benchmarks.

Contribution

The paper proposes a new coordinate-based autoregressive model with dual constraints, addressing limitations of existing diffusion and autoregressive methods in text-to-motion generation.

Findings

Achieves state-of-the-art performance on new benchmarks.

Enhances motion fidelity using diffusion-inspired multi-layer perceptrons.

Guides generation with a dual-constrained causal mask.

Abstract

Text-to-motion generation has attracted increasing attention in the research community recently, with potential applications in animation, virtual reality, robotics, and human-computer interaction. Diffusion and autoregressive models are two popular and parallel research directions for text-to-motion generation. However, diffusion models often suffer from error amplification during noise prediction, while autoregressive models exhibit mode collapse due to motion discretization. To address these limitations, we propose a flexible, high-fidelity, and semantically faithful text-to-motion framework, named Coordinate-based Dual-constrained Autoregressive Motion Generation (CDAMD). With motion coordinates as input, CDAMD follows the autoregressive paradigm and leverages diffusion-inspired multi-layer perceptrons to enhance the fidelity of predicted motions. Furthermore, a Dual-Constrained Causal Mask is introduced to guide autoregressive generation, where motion tokens act as priors and are concatenated with textual encodings. Since there is limited work on coordinate-based motion synthesis, we establish new benchmarks for both text-to-motion generation and motion editing. Experimental results demonstrate that our approach achieves state-of-the-art performance in terms of both fidelity and semantic consistency on these benchmarks.