Temporal Consistency-Aware Text-to-Motion Generation
This work improves text-to-motion generation for applications like animation and robotics by focusing on temporal consistency, though it is incremental as it builds on existing two-stage frameworks.
The paper tackles the problem of generating realistic human motion from text by addressing cross-sequence temporal consistency, which reduces semantic misalignments and improves physical plausibility. The proposed TCA-T2M framework achieves state-of-the-art performance on HumanML3D and KIT-ML benchmarks.
Text-to-Motion (T2M) generation aims to synthesize realistic human motion sequences from natural language descriptions. While two-stage frameworks leveraging discrete motion representations have advanced T2M research, they often neglect cross-sequence temporal consistency, i.e., the shared temporal structures present across different instances of the same action. This leads to semantic misalignments and physically implausible motions. To address this limitation, we propose TCA-T2M, a framework for temporal consistency-aware T2M generation. Our approach introduces a temporal consistency-aware spatial VQ-VAE (TCaS-VQ-VAE) for cross-sequence temporal alignment, coupled with a masked motion transformer for text-conditioned motion generation. Additionally, a kinematic constraint block mitigates discretization artifacts to ensure physical plausibility. Experiments on HumanML3D and KIT-ML benchmarks demonstrate that TCA-T2M achieves state-of-the-art performance, highlighting the importance of temporal consistency in robust and coherent T2M generation.