UniMuMo: Unified Text, Music and Motion Generation

TL;DR

UniMuMo is a versatile multimodal model that unifies text, music, and motion generation using a transformer architecture, aligning unpaired data through rhythmic patterns and fine-tuning pre-trained models for efficient cross-modal synthesis.

Contribution

It introduces a unified transformer framework for simultaneous text, music, and motion generation, leveraging rhythmic alignment and pre-trained models to reduce computational costs.

Findings

Achieves competitive results across all modalities

Successfully unifies music and motion generation tasks

Demonstrates effective cross-modal data alignment

Abstract

We introduce UniMuMo, a unified multimodal model capable of taking arbitrary text, music, and motion data as input conditions to generate outputs across all three modalities. To address the lack of time-synchronized data, we align unpaired music and motion data based on rhythmic patterns to leverage existing large-scale music-only and motion-only datasets. By converting music, motion, and text into token-based representation, our model bridges these modalities through a unified encoder-decoder transformer architecture. To support multiple generation tasks within a single framework, we introduce several architectural improvements. We propose encoding motion with a music codebook, mapping motion into the same feature space as music. We introduce a music-motion parallel generation scheme that unifies all music and motion generation tasks into a single transformer decoder architecture with a single training task of music-motion joint generation. Moreover, the model is designed by fine-tuning existing pre-trained single-modality models, significantly reducing computational demands. Extensive experiments demonstrate that UniMuMo achieves competitive results on all unidirectional generation benchmarks across music, motion, and text modalities. Quantitative results are available in the \href{https://hanyangclarence.github.io/unimumo_demo/}{project page}.