aiqbench analysis
This research introduces G6D, a learning-free 6D pose estimation solver that eliminates reliance on large pretrained models and GPU resources by leveraging purely geometric matching. By enabling CPU-only deployment with interpretable representations, the method addresses critical constraints for resource-limited robotic platforms while maintaining high accuracy on standard datasets. The approach represents a significant step toward efficient, adaptable computer vision systems essential for real-world automation.
G6D: Geometric Learning-Free RGB-D 6D Pose Solver for Robotic Manipulation
arXiv:2609.23566v1 Announce Type: cross Abstract: 6D object pose estimation is fundamental to robotic manipulation and automation. Recent zero-shot methods have significantly improved generalization to unseen objects, but most still rely on large-scale pretrained models with substantial GPU computation and memory demands. These requirements complicate deployment on robotic platforms where perception, planning, and control share limited computational resources, while learned intermediate representations offer limited geometric interpretability for task-specific adaptation. To address these limitations, we propose G6D, a learning-free, geometry-driven RGB-D 6D pose solver. Given an RGB-D observation, an object instance mask, camera intrinsics, and a CAD model, G6D generates pose hypotheses through template-based geometric matching and refines them using silhouette and depth consistency, forming a purely geometry-driven pose estimation paradigm. This paradigm requires neither pretrained visual models nor target-specific training and preserves interpretable geometric representations throughout pose estimation. Moreover, adjustable hypothesis counts provide flexible accuracy-computation trade-offs, while a CPU-only configuration supports deployment without GPU resources. Experiments on LineMOD and five BOP19 datasets demonstrate advanced performance. Real-world pick-and-place experiments further demonstrate G6D's applicability to robotic manipulation. The complete project is publicly available at https://ai4control.github.io/G6D-Project-Page .
Abstract:6D object pose estimation is fundamental to robotic manipulation and automation. Recent zero-shot methods have significantly improved generalization to unseen objects, but most still rely on large-scale pretrained models with substantial GPU computation and memory demands. These requirements complicate deployment on robotic platforms where perception, planning, and control share limited computational resources, while learned intermediate representations offer limited geometric interpretability for task-specific adaptation. To address these limitations, we propose G6D, a learning-free, geometry-driven RGB-D 6D pose solver. Given an RGB-D observation, an object instance mask, camera intrinsics, and a CAD model, G6D generates pose hypotheses through template-based geometric matching and refines them using silhouette and depth consistency, forming a purely geometry-driven pose estimation paradigm. This paradigm requires neither pretrained visual models nor target-specific training and preserves interpretable geometric representations throughout pose estimation. Moreover, adjustable hypothesis counts provide flexible accuracy-computation trade-offs, while a CPU-only configuration supports deployment without GPU resources. Experiments on LineMOD and five BOP19 datasets demonstrate advanced performance. Real-world pick-and-place experiments further demonstrate G6D's applicability to robotic manipulation. The complete project is publicly available at this https URL .
9 pages, 5 figures. Corresponding author: Chuxiong Hu. Project page: this https URL
Computer Vision and Pattern Recognition (cs.CV); Robotics (cs.RO)
arXiv:2609.23566 [cs.CV]
(or arXiv:2609.23566v1 [cs.CV] for this version)
https://doi.org/10.48550/arXiv.2609.23566 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 · 292 words
Related coverage
Real-Time and Accurate Collision-Free Teleoperation via Differentiable Constraint-Based Trajectory Planning
arXiv · Jun 9, 2026
The Machines that Make the Machines
NVIDIA · Oct 7, 2026
GeoLAM: Learning Geometry-Grounded Latent Actions from Unlabeled Human Videos
arXiv · Sep 16, 2026
SemAnCorr: Semantic Anchored Correspondence for Zero-Shot Manipulation Skill Transfer
arXiv · Jul 31, 2026