UniPose9D: Universal Category-Agnostic Object Pose Estimation
2026-07-10 • Computer Vision and Pattern Recognition
Computer Vision and Pattern Recognition
AI summaryⓘ
The authors developed UniPose9D, a model that estimates an object's 3D position, orientation, and size using only images and depth data, without needing labels or templates. It works by analyzing pairs of points on the object's surface and uses advanced feature extraction and a novel RANSAC-based algorithm to improve accuracy. They also handle symmetry problems and combine data from multiple sources to train the model. Tests show that their unified approach performs as well as or better than specialized methods on different datasets and can handle new objects and real-world scenes.
object pose estimation9D poseRGB-D imagingNOCS coordinatesDINOv2PointNetRANSACKabsch algorithmflow matchingsymmetry ambiguity
Authors
Yang You, Yi Du, Cole Harrison, Leonidas Guibas
Abstract
Object pose estimation is a fundamental problem in 3D vision. Although recent state-of-the-art approaches achieve strong performance, they often overfit to existing benchmarks and exhibit limited generalization to novel categories and unseen scenes. We propose UniPose9D, a category-agnostic foundation model for 9D object pose estimation: given an instance mask/ROI and either an RGB-D observation or an RGB image with predicted depth, the model estimates rotation, translation, and metric size without category labels, CAD models, mean-shape priors, or reference views. Specifically, UniPose9D samples point pairs from the observed object geometry and uses DINOv2 and PointNet features to predict NOCS coordinates for each pair. To improve accuracy, we introduce a point-pair-based RANSAC N-hop Kabsch--Umeyama algorithm with an adaptive threshold. We further employ flow matching to address symmetric ambiguities and construct a large-scale training set by curating and aligning pose annotations from existing public datasets. Experiments across six datasets show that a single unified model can match or surpass specialist methods while generalizing to unseen objects and in-the-wild scenarios. Our code and model are available on https://github.com/qq456cvb/UniPose9D.