analiza aiqbench
StrataVLA addresses the critical spatial awareness gap in Vision-Language-Action models by integrating hierarchical geometric grounding without requiring explicit depth inputs during inference. By leveraging frozen geometry foundations and layer-specific adapters, the framework achieves high robotic manipulation success rates while significantly reducing computational overhead through efficient feature caching. This approach demonstrates a viable path toward scalable, spatially grounded autonomous agents that retain the semantic strengths of large-scale pretraining.
StrataVLA: Hierarchical and Efficient 3D Geometric Grounding for Vision-Language-Action Models
arXiv:2609.26071v1 Announce Type: new Abstract: Vision-Language-Action (VLA) models inherit strong semantic priors from large-scale vision-language pretraining, yet remain limited in robotic manipulation by insufficient 3D spatial awareness. Existing approaches either require explicit depth or point-cloud inputs, compress geometry into training-time supervision, or inject it only at the model input or action expert, leaving the vision-language backbone without persistent access to task-relevant spatial information. We introduce StrataVLA, a plug-and-play framework for hierarchical geometric grounding. A frozen geometry foundation model extracts shared geometric features from RGB observations, while sparse, layer-specific Geometry Adapters allow visual representations at selected backbone depths to retrieve relevant geometric evidence through cross-attention. To make inference-time geometry practical, StrataVLA further combines task-aware routing with an LRU feature cache that exploits temporal redundancy during task manipulation. Experiments on LIBERO, SimplerEnv, and real-world manipulation demonstrate consistent gains over strong VLA baselines. StrataVLA achieves 98.53% average success on LIBERO suites while reducing geometry-model invocations by up to 88%, establishing hierarchical geometry injection as an effective and efficient way to achieve spatially grounded robotic control.
Abstract:Vision-Language-Action (VLA) models inherit strong semantic priors from large-scale vision-language pretraining, yet remain limited in robotic manipulation by insufficient 3D spatial awareness. Existing approaches either require explicit depth or point-cloud inputs, compress geometry into training-time supervision, or inject it only at the model input or action expert, leaving the vision-language backbone without persistent access to task-relevant spatial information. We introduce StrataVLA, a plug-and-play framework for hierarchical geometric grounding. A frozen geometry foundation model extracts shared geometric features from RGB observations, while sparse, layer-specific Geometry Adapters allow visual representations at selected backbone depths to retrieve relevant geometric evidence through cross-attention. To make inference-time geometry practical, StrataVLA further combines task-aware routing with an LRU feature cache that exploits temporal redundancy during task manipulation. Experiments on LIBERO, SimplerEnv, and real-world manipulation demonstrate consistent gains over strong VLA baselines. StrataVLA achieves 98.53% average success on LIBERO suites while reducing geometry-model invocations by up to 88%, establishing hierarchical geometry injection as an effective and efficient way to achieve spatially grounded robotic control.
Robotics (cs.RO)
arXiv:2609.26071 [cs.RO]
(or arXiv:2609.26071v1 [cs.RO] for this version)
https://doi.org/10.48550/arXiv.2609.26071 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 260 cuvinte
Subiecte conexe
GroundingVLN: Reasoning and Acting with Grounding for Vision-Language Navigation
arXiv · 17 sept. 2026
GAE: Unleashing Physical Potential of VLM with Generalizable Action Expert
arXiv · 12 iun. 2026
HIGS: Hierarchical Implicit Grids for Joint Geometric and Semantic Scene Understanding
arXiv · 1 oct. 2026
Dyna-2: General Purpose Physical Agents with Hour-Long Autonomy — Jason Ma, Dyna Robotics
Latent Space · 13 oct. 2026