Exploring Motion-Language Alignment for Text-driven Motion Generation

arXiv:2604.02973v2 Announce Type: replace Abstract: Text-driven human motion generation aims to synthesize realistic motion sequences that follow textual descriptions. Despite recent advances, accurately aligning motion dynamics with textual semantics remains a fundamental challenge. In this paper, we revisit text-to-motion generation from the perspective of motion-language alignment and propose MLA-Gen, a framework that integrates global motion priors with fine-grained local conditioning. This design enables the model to capture common motion patterns, while establishing detailed alignment between texts and motions. Furthermore, we identify a previously overlooked attention sink phenomenon in human motion generation, where attention disproportionately concentrates on the start text token, limiting the utilization of informative textual cues and leading to degraded semantic grounding. To analyze this issue, we introduce SinkRatio, a metric for measuring attention concentration, and develop alignment-aware masking and control strategies to regulate attention during generation. Extensive experiments demonstrate that our approach consistently improves both motion quality and motion-language alignment over strong baselines.

Related

• 3DAlign-DAER: Dynamic Attention Policy and Efficient Retrieval Strategy for Fine-grained 3D-Text Alignment at Scale
• UniMoFlow: Grounding Instruction-Driven 3D Human Motion Editing in Generation
• IAM: Identity-Aware Human Motion and Shape Joint Generation
• VibeAvatar: Aligning Phonetic Kinematics and Human Aesthetics for High-Fidelity Talking Avatar Synthesis
• Track the Noise, Move the World:3D-Grounded Motion-Consistent Noise for Controllable Video Generation

Source: arXiv cs.CV | 2026-10-14