The Quest for Generalizable Motion Generation: Data, Model, and Evaluation

TL;DR

This paper introduces a comprehensive framework for improving generalization in 3D human motion generation by leveraging insights from video generation, including a large dataset, a novel diffusion model, and a new benchmark.

Contribution

It presents ViMoGen, a flow-matching diffusion transformer that integrates data, modeling, and evaluation strategies from video generation to enhance motion generation.

Findings

The new dataset ViMoGen-228K significantly expands semantic diversity.

The proposed ViMoGen model outperforms existing methods in quality and generalization.

The benchmark MBench enables fine-grained evaluation of motion generation models.

Abstract

Despite recent advances in 3D human motion generation (MoGen) on standard benchmarks, existing text-to-motion models still face a fundamental bottleneck in their generalization capability. In contrast, adjacent generative fields, most notably video generation (ViGen), have demonstrated remarkable generalization in modeling human behaviors, highlighting transferable insights that MoGen can leverage. Motivated by this observation, we present a comprehensive framework that systematically transfers knowledge from ViGen to MoGen across three key pillars: data, modeling, and evaluation. First, we introduce ViMoGen-228K, a large-scale dataset comprising 228,000 high-quality motion samples that integrates high-fidelity optical MoCap data with semantically annotated motions from web videos and synthesized samples generated by state-of-the-art ViGen models. The dataset includes both text-motion pairs and text-video-motion triplets, substantially expanding semantic diversity. Second, we propose ViMoGen, a flow-matching-based diffusion transformer that unifies priors from MoCap data and ViGen models through gated multimodal conditioning. To enhance efficiency, we further develop ViMoGen-light, a distilled variant that eliminates video generation dependencies while preserving strong generalization. Finally, we present MBench, a hierarchical benchmark designed for fine-grained evaluation across motion quality, prompt fidelity, and generalization ability. Extensive experiments show that our framework significantly outperforms existing approaches in both automatic and human evaluations. The code, data, and benchmark will be made publicly available. Homepage: https://motrixlab.github.io/2026_iclr_vimogen.