T2M Mamba: Motion Periodicity-Saliency Coupling Approach for Stable Text-Driven Motion Generation
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2026
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866910007806656512 |
|---|---|
| author | Zhan, Xingzu Xie, Chen Chen, Honghang Lin, Yixun Mai, Xiaochun |
| author_facet | Zhan, Xingzu Xie, Chen Chen, Honghang Lin, Yixun Mai, Xiaochun |
| contents | Text-to-motion generation, which converts motion language descriptions into coherent 3D human motion sequences, has attracted increasing attention in fields, such as avatar animation and humanoid robotic interaction. Though existing models have achieved significant fidelity, they still suffer from two core limitations: (i) They treat motion periodicity and keyframe saliency as independent factors, overlooking their coupling and causing generation drift in long sequences. (ii) They are fragile to semantically equivalent paraphrases, where minor synonym substitutions distort textual embeddings, propagating through the decoder and producing unstable or erroneous motions. In this work, we propose T2M Mamba to address these limitations by (i) proposing Periodicity-Saliency Aware Mamba, which utilizes novel algorithms for keyframe weight estimation via enhanced Density Peaks Clustering and motion periodicity estimation via FFT-accelerated autocorrelation to capture coupled dynamics with minimal computational overhead, and (ii) constructing a Periodic Differential Cross-modal Alignment Module (PDCAM) to enhance robust alignment of textual and motion embeddings. Extensive experiments on HumanML3D and KIT-ML datasets have been conducted, confirming the effectiveness of our approach, achieving an FID of 0.068 and consistent gains on all other metrics. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_01352 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | T2M Mamba: Motion Periodicity-Saliency Coupling Approach for Stable Text-Driven Motion Generation Zhan, Xingzu Xie, Chen Chen, Honghang Lin, Yixun Mai, Xiaochun Computer Vision and Pattern Recognition Text-to-motion generation, which converts motion language descriptions into coherent 3D human motion sequences, has attracted increasing attention in fields, such as avatar animation and humanoid robotic interaction. Though existing models have achieved significant fidelity, they still suffer from two core limitations: (i) They treat motion periodicity and keyframe saliency as independent factors, overlooking their coupling and causing generation drift in long sequences. (ii) They are fragile to semantically equivalent paraphrases, where minor synonym substitutions distort textual embeddings, propagating through the decoder and producing unstable or erroneous motions. In this work, we propose T2M Mamba to address these limitations by (i) proposing Periodicity-Saliency Aware Mamba, which utilizes novel algorithms for keyframe weight estimation via enhanced Density Peaks Clustering and motion periodicity estimation via FFT-accelerated autocorrelation to capture coupled dynamics with minimal computational overhead, and (ii) constructing a Periodic Differential Cross-modal Alignment Module (PDCAM) to enhance robust alignment of textual and motion embeddings. Extensive experiments on HumanML3D and KIT-ML datasets have been conducted, confirming the effectiveness of our approach, achieving an FID of 0.068 and consistent gains on all other metrics. |
| title | T2M Mamba: Motion Periodicity-Saliency Coupling Approach for Stable Text-Driven Motion Generation |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2602.01352 |