aiqbench analysis
This research addresses the reliability challenges of diffusion-based Vision-Language-Action (VLA) policies by introducing Selected Diffusion Noise (SDN), a training-free technique that enhances test-time robustness against spurious visual inputs. By filtering action candidates based on visual grounding and kinematic stability metrics like JerkRMS, the method significantly reduces stochastic failures in both simulated and real-world robotic manipulation tasks.
Test-Time Improvement of VLA Policies via Selected Diffusion Noise for Spurious-Robust Action Smoothing
arXiv:2606.14084v2 Announce Type: replace Abstract: Diffusion-based Vision-Language-Action (VLA) policies have demonstrated strong generalization in robot manipulation, but their deployment can remain unreliable due to spurious visual reliance and stochastic action generation, leading to grounding failures and kinematically unstable behaviors. We introduce Selected Diffusion Noise (SDN), a training-free method that treats diffusion noise as a controllable degree of freedom for improving test-time reliability. SDN samples multiple candidate action chunks and applies two-stage selection: it first filters candidates based on task-relevant visual grounding using a counterfactual negative set, then selects the kinematically most stable chunk based on JerkRMS. Experiments on three VLA policies across both simulation and real-world environments show improvements of 8-10% and 18.33%, respectively, demonstrating SDN's effectiveness for test-time improvement.
Authors:Duc Minh Nguyen, Bao-Ngoc Dao, Tung M. Luu, Binh Gia Nguyen, Vinh Tong, Anji Liu, Vu N. Duong, Dung D. Le, Daniel Sonntag, Trung Le, Ngan Le, Jan Peter, An Thai Le, Minh Nhat Vu, Mathias Niepert, Khoa D. Doan, Duy M. H. Nguyen, Vien Anh Ngo
Abstract:Diffusion-based Vision-Language-Action (VLA) policies have demonstrated strong generalization in robot manipulation, but their deployment can remain unreliable due to spurious visual reliance and stochastic action generation, leading to grounding failures and kinematically unstable behaviors. We introduce Selected Diffusion Noise (SDN), a training-free method that treats diffusion noise as a controllable degree of freedom for improving test-time reliability. SDN samples multiple candidate action chunks and applies two-stage selection: it first filters candidates based on task-relevant visual grounding using a counterfactual negative set, then selects the kinematically most stable chunk based on JerkRMS. Experiments on three VLA policies across both simulation and real-world environments show improvements of 8-10% and 18.33%, respectively, demonstrating SDN's effectiveness for test-time improvement.
Robotics (cs.RO)
arXiv:2606.14084 [cs.RO]
(or arXiv:2606.14084v2 [cs.RO] for this version)
Full text under the publisher's open licence · 172 words