# Category-level 6D Object Pose Recovery in Depth Images

arXiv: https://arxiv.org/abs/1808.00255 · HTML: https://mlsanity.com/paper/1808.00255/

## Key facts

- Cited by 21 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 1 Aug 2018.

## Built on by (most cited first, 21 of 21)

- [GPV-Pose: Category-level Object Pose Estimation via Geometry-guided Point-wise Voting](https://mlsanity.com/paper/2203.07918/) (2022)
- [Category-Level 6D Object Pose Estimation via Cascaded Relation and Recurrent Reconstruction Networks](https://mlsanity.com/paper/2108.08755/) (2021)
- [Zero-Shot Category-Level Object Pose Estimation](https://mlsanity.com/paper/2204.03635/) (2022)
- [SAR-Net: Shape Alignment and Recovery Network for Category-level 6D Object Pose and Size Estimation](https://mlsanity.com/paper/2106.14193/) (2021)
- [HS-Pose: Hybrid Scope Feature Extraction for Category-level Object Pose Estimation](https://mlsanity.com/paper/2303.15743/) (2023)
- [IST-Net: Prior-free Category-level Pose Estimation with Implicit Space Transformation](https://mlsanity.com/paper/2303.13479/) (2023)
- [Deep Learning-Based Object Pose Estimation: A Comprehensive Survey](https://mlsanity.com/paper/2405.07801/) (2024)
- [Deep Generative Models in Robotics: A Survey on Learning from Multimodal Demonstrations](https://mlsanity.com/paper/2408.04380/) (2024)
- [Deep Learning on Monocular Object Pose Detection and Tracking: A Comprehensive Overview](https://mlsanity.com/paper/2105.14291/) (2021)
- [USEEK: Unsupervised SE(3)-Equivariant 3D Keypoints for Generalizable Manipulation](https://mlsanity.com/paper/2209.13864/) (2022)
- [StereoPose: Category-Level 6D Transparent Object Pose Estimation from Stereo Images via Back-View NOCS](https://mlsanity.com/paper/2211.01644/) (2022)
- [CHORD: Category-level Hand-held Object Reconstruction via Shape Deformation](https://mlsanity.com/paper/2308.10574/) (2023)
- [A Visual Navigation Perspective for Category-Level Object Pose Estimation](https://mlsanity.com/paper/2203.13572/) (2022)
- [You Only Look at One: Category-Level Object Representations for Pose Estimation From a Single Example](https://mlsanity.com/paper/2305.12626/) (2023)
- [Optimal and Robust Category-level Perception: Object Pose and Shape Estimation from 2D and 3D Semantic Keypoints](https://mlsanity.com/paper/2206.12498/) (2022)
- [Synthetic data enables faster annotation and robust segmentation for multi-object grasping in clutter](https://mlsanity.com/paper/2401.13405/) (2024)
- [Joint Learning of Pose Regression and Denoising Diffusion with Score Scaling Sampling for Category-level 6D Pose Estimation](https://mlsanity.com/paper/2510.04125/) (2025)
- [FastPoseCNN: Real-Time Monocular Category-Level Pose and Size Estimation Framework](https://mlsanity.com/paper/2406.11063/) (2024)
- [Universal Features Guided Zero-Shot Category-Level Object Pose Estimation](https://mlsanity.com/paper/2501.02831/) (2025)
- [ImaginationPolicy: Towards Generalizable, Precise and Reliable End-to-End Policy for Robotic Manipulation](https://mlsanity.com/paper/2509.20841/) (2025)
- [RFM-Pose:Reinforcement-Guided Flow Matching for Fast Category-Level 6D Pose Estimation](https://mlsanity.com/paper/2602.05257/) (2026)

## Abstract (arXiv)

Intra-class variations, distribution shifts among source and target domains are the major challenges of category-level tasks. In this study, we address category-level full 6D object pose estimation in the context of depth modality, introducing a novel part-based architecture that can tackle the above-mentioned challenges. Our architecture particularly adapts the distribution shifts arising from shape discrepancies, and naturally removes the variations of texture, illumination, pose, etc., so we call it as "Intrinsic Structure Adaptor (ISA)". We engineer ISA based on the followings: i) "Semantically Selected Centers (SSC)" are proposed in order to define the "6D pose" at the level of categories. ii) 3D skeleton structures, which we derive as shape-invariant features, are used to represent the parts extracted from the instances of given categories, and privileged one-class learning is employed based on these parts. iii) Graph matching is performed during training in such a way that the adaptation/generalization capability of the proposed architecture is improved across unseen instances. Experiments validate the promising performance of the proposed architecture on both synthetic and real datasets.