arXiv AI

AnyBox: Efficient Zero-Shot 9DoF Pose Estimation of Boxes for Robotic Manipulation

AnyBox is a zero‑shot framework that estimates the full 9DoF pose (6D pose plus 3D dimensions) of boxes from a single RGB‑D image, leveraging the geometric regularity of boxes. It alternates between pose and scale estimation, using a binary search guided by the discrepancy between a reprojected template and the observed mask, and employs a depth‑consistency filter and an early‑stopping rule to prune implausible hypotheses. On public benchmarks and a warehouse dataset, AnyBox improves detection AP by up to 36 points and boosts robotic box‑shelving success by 28%.

arXiv Computer Vision
2d ago

PriorPose: Reference-Guided Joint Deformation and Alignment for Category-Level Object Pose Estimation

PriorPose introduces a reference-guided correspondence framework for category-level object pose estimation that jointly performs canonicalization and alignment in a shared feature space. By embedding partial observations and a category prior as token sets in a seeded transformer, the network predicts per-point NOCS fields and a canonical deformation, while a deep pose head regresses the similarity transform. A two-part shape consistency objective couples correspondence, deformation, and pose, reducing reliance on memorized canonical orientations and avoiding error cascades, leading to state-of-the-art results on standard and larger-category benchmarks, especially under strict pose thresholds.

By Yihan Chen, Huan Ren, Wenfei Yang, Hang Du, Tianzhu Zhang, Feng Wu
arXiv Computer Vision
Aug 24

From Simulation to the Real-World: An In-Field 6D Pose Dataset and Baseline for Robotic Strawberry Harvesting

The paper presents the first real‑world 6D pose ground‑truth dataset for red‑stage strawberries, collected from 12,040 images at an actual farm using indirect camera pose recovery and 3D bounding‑box annotation. It also introduces a synthetic dataset rendered in NVIDIA Isaac Sim with scene‑level realism and domain randomization. Experiments show that models trained solely on synthetic data do not transfer well to in‑field images, but adding a small amount of real data significantly improves both translation and rotation accuracy across various backbone encoders.

By Woojung Son (Department of Agricultural and Biological Engineering, University of Florida), Won Suk Lee (Department of Agricultural and Biological Engineering, University of Florida), Zijing Huang (Department of Agricultural and Biological Engineering, University of Florida), Daeun Choi (Department of Agricultural and Biological Engineering, University of Florida), Catia Silva (Department of Electrical and Computer Engineering, University of Florida), Yu She (Edwardson School of Industrial Engineering, Purdue University), Yan Gu (School of Mechanical Engineering, Purdue University)
arXiv AI
Sep 7

Where Appearance Fails, Geometry Recognizes: A CAD-Free 3D Shape Prior That Complements Vision Foundation Models

The paper introduces a CAD‑free 3D shape prior that enhances object recognition by reconstructing each object with 3D Gaussian Splatting (3DGS) from short RGB‑D scans and fusing the resulting shape prototype with frozen DINOv2 image features. Experiments on T‑LESS and HOPE datasets show that geometry alone can match or exceed CAD‑based recognition, and that the combined approach improves performance, especially on shape‑distinctive or partially occluded objects. The study demonstrates that the benefit comes from the geometric information rather than rendered pixels, and that the prior is complementary to frozen vision features.

By Chenxi Tao, Seung-Kyum Choi
arXiv Computer Vision
Aug 31

SUFLECA: Scaling Up Feature Learning for CAD-to-image Alignment

SUFLECA is a weakly supervised framework that improves zero‑shot CAD‑to‑image alignment by scaling geometry‑grounded feature learning using Normalized Object Coordinates across up to 12 real and synthetic datasets. It introduces a geometrically consistent matching algorithm that reliably establishes CAD‑to‑image correspondences, enabling accurate, sub‑second alignment without iterative pose refinement. On the ScanNet25k benchmark, SUFLECA achieves 32.8%/42.6% category/instance accuracy, outperforming the strongest zero‑shot baseline by 9.7/12.5 percentage points and surpassing existing pose‑supervised methods for the first time.

By Saad Ejaz, Miguel Fernandez-Cortizas, Javier Civera, Holger Voos, Jose Luis Sanchez-Lopez
arXiv AI
Jul 29

SAM3D-Guided Object-Centric Representation Alignment for Vision-Language-Action Models

arXiv:2607. 25912v1 Announce Type: cross Abstract: Vision-Language-Action (VLA) models have shown strong potential for general robot manipulation, but most existing models rely on 2D visual-language backbones and lack fine-grained 3D understanding of target objects, especially under occlusion, pose variation, scale changes, and precise spatial interaction.

By Zonghe Liu (University of Hong Kong), Shanyuan Jie (Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences), Xiaoquan Sun (Huazhong University of Science and Technology), Chen Cao (University of Hong Kong), Zetian Xu (University of Hong Kong), Zongsheng Liu (Beijing University of Aeronautics and Astronautics), Jiayu Chen (University of Hong Kong, Infiforce)
arXiv Machine Learning
Jun 19

Pose6DAug: Physically Plausible Multi-view Object Swapping for Robot Data Augmentation

arXiv:2606. 20118v1 Announce Type: cross Abstract: Vision-language-action (VLA) policies have shown strong potential for general-purpose manipulation, yet they often fail on novel, out-of-distribution objects whose appearance or geometry deviates from the training distribution.

By Jonghoon Lee, Seong Hyeon Park, Byungwoo Jeon, Minha Lee, Jinwoo Shin
arXiv Computer Vision
1d ago

CALIPER: Metric-Grounded Model-Free Recognition of Visually Similar Industrial Parts

CALIPER is a model‑free RGB‑D framework that performs fine‑grained recognition of visually similar industrial parts by combining support‑based appearance matching with metric size evidence. Each class is onboarded from a single turntable RGB‑D video and a few labeled real images, enabling 3D reconstruction for appearance support and depth‑aligned size profiling. At inference, a YOLOv8n‑seg model localizes parts, a frozen DINOv2 backbone with an episodically trained embedding head matches support, and margin‑conditioned metric fusion selectively uses size evidence for ambiguous cases, achieving high accuracy on 18 parts and robust enrollment of unseen screws without retraining.

By Alankrit Gupta, Chenxi Tao, Seung-Kyum Choi
arXiv Computer Vision
Aug 27

Point2Pose: Occlusion-Recovering 6D Pose Tracking and 3D Reconstruction for Multiple Unknown Objects Via 2D Point Trackers

Point2Pose is a model‑free method for causal 6D pose tracking of multiple rigid objects using monocular RGB‑D video. It starts from sparse image points and employs a 2D point tracker to maintain long‑range correspondences, allowing instant recovery after complete occlusion. The system also incrementally builds an online Truncated Signed Distance Function (TSDF) representation of the tracked objects and introduces a new multi‑object tracking dataset with motion‑capture ground truth.

By Tzu-Yuan Lin, Ho Jae Lee, Kevin Doherty, Yonghyeon Lee, Sangbae Kim
arXiv Computer Vision
Sep 3

MV-dVRK: A Multi-Viewpoint Benchmark for Spatial Surgical Perception

MV-dVRK is the first ex‑vivo surgical dataset that provides multiple exposure‑synchronized stereo viewpoints, accurate surface geometry, and ground‑truth camera poses for endoscopic images. The benchmark’s static subset offers dense SfM reference geometry validated against an industrial 3D scanner, while the dynamic sequences cover ten surgical tasks with increasing kinematic complexity and tissue deformation. Using MV‑dVRK, the authors systematically compare zero‑shot monocular, stereo, multi‑stereo, and multi‑view 3D reconstruction methods, finding that multi‑stereo reconstruction with two endoscopes yields the highest coverage, and that optimization‑based multi‑view methods outperform feed‑forward foundation models when a third viewpoint is added.

By Guido Caccianiga, Sergey Prokudin, Yutong Chen, Bernard Javot, Rachael L'Orsa, Omer Burak Alada\u{g}, Yarden Sharon, Jens Rolinger, Ivan Capobianco, Anton Deguet, Siyu Tang, Katherine J. Kuchenbecker