├── README.md └── teaser ├── robot.gif └── teaser.png /README.md: -------------------------------------------------------------------------------- 1 | # Any6D: Model-free 6D Pose Estimation of Novel Objects 2 | 3 | This is the official implementation of our paper accepted by CVPR 2025 4 | 5 | [[Website]](https://sites.google.com/view/taeyeop-lee/any6d) [[Paper]](https://arxiv.org/pdf/2503.18673) 6 | 7 | Authors: Taeyeop Lee, Bowen Wen, Minjun Kang, Gyuree Kang, In So Kweon, Kuk-Jin Yoon 8 | 9 |

10 | 11 |

12 | 13 | # Abstract 14 | 15 | We introduce Any6D, a model-free framework for 6D object pose estimation that requires only a single RGB-D anchor image to estimate both the 6D pose and size of unknown objects in novel scenes. Unlike existing methods that rely on textured 3D models or multiple viewpoints, Any6D leverages a joint object alignment process to enhance 2D-3D alignment and metric scale estimation for improved pose accuracy. Our approach integrates a render-and-compare strategy to generate and refine pose hypotheses, enabling robust performance in scenarios with occlusions, non-overlapping views, diverse lighting conditions, and large cross-environment variations. We evaluate our method on five challenging datasets: REAL275, Toyota-Light, HO3D, YCBINEOAT, and LM-O, demonstrating its effectiveness in significantly outperforming state-of-the-art methods for novel object pose estimation. 16 | 17 | 18 |

19 | 20 |

21 | 22 | 23 | 24 | # Coming Soon (March ~ May 2025) 25 | -------------------------------------------------------------------------------- /teaser/robot.gif: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/taeyeopl/Any6D/94f432af2b4f75d865ac54f21ac5b03909866c81/teaser/robot.gif -------------------------------------------------------------------------------- /teaser/teaser.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/taeyeopl/Any6D/94f432af2b4f75d865ac54f21ac5b03909866c81/teaser/teaser.png --------------------------------------------------------------------------------