Dyadic Mamba: Long-term Dyadic Human Motion Synthesis

TL;DR

Dyadic Mamba introduces a novel SSM-based method for long-term dyadic human motion synthesis from text, overcoming transformer limitations and establishing a new benchmark for extended interaction quality.

Contribution

The paper presents Dyadic Mamba, a simple SSM-based architecture that effectively generates long-term dyadic human motion, outperforming transformers on extended sequences and introducing a new evaluation benchmark.

Findings

Achieves competitive short-term performance

Significantly outperforms transformers on long sequences

Proposes a new benchmark for long-term motion synthesis

Abstract

Generating realistic dyadic human motion from text descriptions presents significant challenges, particularly for extended interactions that exceed typical training sequence lengths. While recent transformer-based approaches have shown promising results for short-term dyadic motion synthesis, they struggle with longer sequences due to inherent limitations in positional encoding schemes. In this paper, we introduce Dyadic Mamba, a novel approach that leverages State-Space Models (SSMs) to generate high-quality dyadic human motion of arbitrary length. Our method employs a simple yet effective architecture that facilitates information flow between individual motion sequences through concatenation, eliminating the need for complex cross-attention mechanisms. We demonstrate that Dyadic Mamba achieves competitive performance on standard short-term benchmarks while significantly outperforming transformer-based approaches on longer sequences. Additionally, we propose a new benchmark for evaluating long-term motion synthesis quality, providing a standardized framework for future research. Our results demonstrate that SSM-based architectures offer a promising direction for addressing the challenging task of long-term dyadic human motion synthesis from text descriptions.