DyQ-VLA: Temporal-Dynamic-Aware Quantization for Embodied Vision-Language-Action Models
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2026
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866917340316172288 |
|---|---|
| author | Zheng, Zihao Cao, Hangyu Tian, Sicheng Chen, Jiayu Li, Maoliang Sun, Xinhao Zou, Hailong Zhang, Zhaobo Liu, Xuanzhe Cao, Donggang Mei, Hong Chen, Xiang |
| author_facet | Zheng, Zihao Cao, Hangyu Tian, Sicheng Chen, Jiayu Li, Maoliang Sun, Xinhao Zou, Hailong Zhang, Zhaobo Liu, Xuanzhe Cao, Donggang Mei, Hong Chen, Xiang |
| contents | Vision-Language-Action (VLA) models are dominant in embodied intelligence but are constrained by inference overheads. While model quantization alleviates these bottlenecks for edge deployment, static quantization approaches remain suboptimal for VLAs due to two critical challenges: (1) Temporal-dynamic sensitivity, where fixed precision wastes resources by ignoring stage-varying error tolerances; and (2) Real-time allocation, where identifying real-time sensitivity to guide bit allocation remains unsolved. To address these challenges, we propose DyQ-VLA, a dynamic quantization framework for VLAs. Specifically, a sensitivity-aware switching strategy leverages real-time kinematic proxies to trigger the bit-width switch, while a kinematic-guided module dynamically allocates the optimal bit-width. Experiments show that DyQ-VLA requires only 30.9% of the original memory footprint while maintaining 99.5% of its original performance, achieving 1.49x simulation and up to 1.43x real-world speedups. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_07904 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | DyQ-VLA: Temporal-Dynamic-Aware Quantization for Embodied Vision-Language-Action Models Zheng, Zihao Cao, Hangyu Tian, Sicheng Chen, Jiayu Li, Maoliang Sun, Xinhao Zou, Hailong Zhang, Zhaobo Liu, Xuanzhe Cao, Donggang Mei, Hong Chen, Xiang Machine Learning Robotics Vision-Language-Action (VLA) models are dominant in embodied intelligence but are constrained by inference overheads. While model quantization alleviates these bottlenecks for edge deployment, static quantization approaches remain suboptimal for VLAs due to two critical challenges: (1) Temporal-dynamic sensitivity, where fixed precision wastes resources by ignoring stage-varying error tolerances; and (2) Real-time allocation, where identifying real-time sensitivity to guide bit allocation remains unsolved. To address these challenges, we propose DyQ-VLA, a dynamic quantization framework for VLAs. Specifically, a sensitivity-aware switching strategy leverages real-time kinematic proxies to trigger the bit-width switch, while a kinematic-guided module dynamically allocates the optimal bit-width. Experiments show that DyQ-VLA requires only 30.9% of the original memory footprint while maintaining 99.5% of its original performance, achieving 1.49x simulation and up to 1.43x real-world speedups. |
| title | DyQ-VLA: Temporal-Dynamic-Aware Quantization for Embodied Vision-Language-Action Models |
| topic | Machine Learning Robotics |
| url | https://arxiv.org/abs/2603.07904 |