ACG: Action Coherence Guidance for Flow-based Vision-Language-Action models
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914420788035584 |
|---|---|
| author | Park, Minho Kim, Kinam Hyung, Junha Jang, Hyojin Jin, Hoiyeong Yun, Jooyeol Lee, Hojoon Choo, Jaegul |
| author_facet | Park, Minho Kim, Kinam Hyung, Junha Jang, Hyojin Jin, Hoiyeong Yun, Jooyeol Lee, Hojoon Choo, Jaegul |
| contents | Diffusion and flow matching models have emerged as powerful robot policies, enabling Vision-Language-Action (VLA) models to generalize across diverse scenes and instructions. Yet, when trained via imitation learning, their high generative capacity makes them sensitive to noise in human demonstrations: jerks, pauses, and jitter which reduce action coherence. Reduced action coherence causes instability and trajectory drift during deployment, failures that are catastrophic in fine-grained manipulation where precision is crucial. In this paper, we present Action Coherence Guidance (ACG) for VLA models, a training-free test-time guidance algorithm that improves action coherence and thereby yields performance gains. Evaluated on RoboCasa, DexMimicGen, and real-world SO-101 tasks, ACG consistently improves action coherence and boosts success rates across diverse manipulation tasks. Code and project page are available at https://github.com/DAVIAN-Robotics/ACG and https://DAVIAN-Robotics.github.io/ACG , respectively. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_22201 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | ACG: Action Coherence Guidance for Flow-based Vision-Language-Action models Park, Minho Kim, Kinam Hyung, Junha Jang, Hyojin Jin, Hoiyeong Yun, Jooyeol Lee, Hojoon Choo, Jaegul Robotics Diffusion and flow matching models have emerged as powerful robot policies, enabling Vision-Language-Action (VLA) models to generalize across diverse scenes and instructions. Yet, when trained via imitation learning, their high generative capacity makes them sensitive to noise in human demonstrations: jerks, pauses, and jitter which reduce action coherence. Reduced action coherence causes instability and trajectory drift during deployment, failures that are catastrophic in fine-grained manipulation where precision is crucial. In this paper, we present Action Coherence Guidance (ACG) for VLA models, a training-free test-time guidance algorithm that improves action coherence and thereby yields performance gains. Evaluated on RoboCasa, DexMimicGen, and real-world SO-101 tasks, ACG consistently improves action coherence and boosts success rates across diverse manipulation tasks. Code and project page are available at https://github.com/DAVIAN-Robotics/ACG and https://DAVIAN-Robotics.github.io/ACG , respectively. |
| title | ACG: Action Coherence Guidance for Flow-based Vision-Language-Action models |
| topic | Robotics |
| url | https://arxiv.org/abs/2510.22201 |