aiqbench analysis
GraspFoM introduces a unified framework that treats 3D reconstruction not as an intermediate step but as a reusable prior to enhance robotic grasping under partial observations. By leveraging shared 3D object latents and foundation models like SAM3D, the method achieves state-of-the-art performance in both tasks while requiring only a minimal increase in trainable parameters. This approach effectively bridges geometric fidelity with grasp affordance prediction, offering a scalable path for next-generation manipulation systems.
GraspFoM: Towards Reconstruction-Driven Robotic Grasping with 3D Foundation Priors
arXiv:2606.08440v1 Announce Type: new Abstract: Robotic grasping is a fundamental capability in robotic manipulation. Yet grasping remains challenging under partial observations. Reliable grasping depends on both local contact cues and object-level 3D structure. Existing geometry-aware grasping methods recognize the value of reconstruction, but they typically treat geometry as an intermediate prediction rather than a reusable object prior for grasping. In this paper, we present GraspFoM, a unified framework that leverages 3D foundation priors (SAM3D) to build a shared 3D object latent for both reconstruction and grasp pose prediction. Built on this shared object latent, we introduce an anchor-initialized truncated pose-reasoning diffuser that predicts continuous and multimodal grasp poses without directly relying on discrete grasp candidates. We further investigate the interaction between reconstruction and grasping through a reconstruction-aware scorer and a residual latent updater. Reconstruction provides grounded geometric cues, while grasp supervision refines the shared object latent toward grasp-relevant affordances. GraspFoM jointly predicts grasp poses and reconstructs high-fidelity 3D assets in mesh and 3DGS forms. Comprehensive experiments demonstrate that GraspFoM achieves state-of-the-art results on both reconstruction and grasping. Notably, these improvements require only a small number of additional trainable parameters. Component-wise ablation studies also demonstrate the contribution of each component.
Abstract:Robotic grasping is a fundamental capability in robotic manipulation. Yet grasping remains challenging under partial observations. Reliable grasping depends on both local contact cues and object-level 3D structure. Existing geometry-aware grasping methods recognize the value of reconstruction, but they typically treat geometry as an intermediate prediction rather than a reusable object prior for grasping. In this paper, we present GraspFoM, a unified framework that leverages 3D foundation priors (SAM3D) to build a shared 3D object latent for both reconstruction and grasp pose prediction. Built on this shared object latent, we introduce an anchor-initialized truncated pose-reasoning diffuser that predicts continuous and multimodal grasp poses without directly relying on discrete grasp candidates. We further investigate the interaction between reconstruction and grasping through a reconstruction-aware scorer and a residual latent updater. Reconstruction provides grounded geometric cues, while grasp supervision refines the shared object latent toward grasp-relevant affordances. GraspFoM jointly predicts grasp poses and reconstructs high-fidelity 3D assets in mesh and 3DGS forms. Comprehensive experiments demonstrate that GraspFoM achieves state-of-the-art results on both reconstruction and grasping. Notably, these improvements require only a small number of additional trainable parameters. Component-wise ablation studies also demonstrate the contribution of each component.
Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV)
arXiv:2606.08440 [cs.RO]
(or arXiv:2606.08440v1 [cs.RO] for this version)
https://doi.org/10.48550/arXiv.2606.08440 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.
Full text under the publisher's open licence · 293 words
Related coverage
The Machines that Make the Machines
NVIDIA · Oct 7, 2026
Dyna-2: General Purpose Physical Agents with Hour-Long Autonomy — Jason Ma, Dyna Robotics
Latent Space · Oct 13, 2026
News Archive 2026
Quantum Computing Report · Jul 25, 2026
Building a Foundation Stack for General-Purpose Robots
IEEE Spectrum · Jul 13, 2026