【论文列表】2025年Object Pose Estimation顶会论文列表

发表于 分类于 阅读次数： Waline： 本文字数： 2.2k 阅读时长 ≈ 8 分钟
整理的2025年Object Pose Estimation顶会论文列表。

CVPR

  1. Any6D: Model-free 6D Pose Estimation of Novel Object
    • 任意物体方法
    • 论文：https://cvpr.thecvf.com/virtual/2025/poster/32502
    • 代码：https://github.com/taeyeopl/Any6D
  2. CAP-Net: A Unified Network for 6D Pose and Size Estimation of Categorical Articulated Parts from a Single RGB-D Image
    • 类别级铰接物体方法
    • 论文：https://cvpr.thecvf.com/virtual/2025/poster/34668
    • 代码：https://github.com/ShaneHuangHZ/CAPNet
  3. Co-op: Correspondence-based Novel Object Pose Estimation
    • 任意物体方法
    • 论文：https://cvpr.thecvf.com/virtual/2025/poster/35005
    • 代码：代码未开源
  4. CRISP: Object Pose and Shape Estimation with Test-Time Adaptation
    • 任意物体方法
    • 论文：https://cvpr.thecvf.com/virtual/2025/poster/35220
    • 代码：https://github.com/MIT-SPARK/CRISP
  5. GCE-Pose: Global Context Enhancement for Category-level Object Pose Estimation
    • 类别级方法
    • 论文：https://cvpr.thecvf.com/virtual/2025/poster/35180
    • 代码：代码未开源
  6. GIVEPose: Gradual Intra-class Variation Elimination for RGB-based Category-Level Object Pose Estimation
    • 类别级方法
    • 论文：https://cvpr.thecvf.com/virtual/2025/poster/33313
    • 代码：https://github.com/ziqin-h/GIVEPose
  7. iG-6DoF: Model-free 6DoF Pose Estimation for Unseen Object via Iterative 3D Gaussian Splatting
    • 任意物体方法
    • 论文：https://cvpr.thecvf.com/virtual/2025/poster/32800
    • 代码：代码未开源
  8. Leveraging Global Stereo Consistency for Category-Level Shape and 6D Pose Estimation from Stereo Images
    • 类别级方法
    • 论文：https://cvpr.thecvf.com/virtual/2025/poster/33782
    • 代码：代码未开源
  9. ONDA-Pose: Occlusion-Aware Neural Domain Adaptation for Self-Supervised 6D Object Pose Estimation
    • 实例级方法
    • 论文：https://cvpr.thecvf.com/virtual/2025/poster/33240
    • 代码：https://github.com/tan-tao11/ONDA-Pose
  10. One2Any: One-Reference 6D Pose Estimation for Any Object
    • 任意物体方法
    • 论文：https://cvpr.thecvf.com/virtual/2025/poster/32415
    • 代码：https://github.com/lmy1001/One2Any
  11. Pos3R: 6D Pose Estimation for Unseen Objects Made Easy
    • 任意物体方法
    • 论文：https://cvpr.thecvf.com/virtual/2025/poster/35115
    • 代码：代码未开源
  12. RefPose: Leveraging Reference Geometric Correspondences for Accurate 6D Pose Estimation of Unseen Objects
    • 任意物体方法
    • 论文：https://cvpr.thecvf.com/virtual/2025/poster/34376
    • 代码：代码未开源
  13. Rethinking Correspondence-based Category-Level Object Pose Estimation
    • 类别级方法
    • 论文：https://cvpr.thecvf.com/virtual/2025/poster/34619
    • 代码：代码未开源
  14. SCFlow2: Plug-and-Play Object Pose Refiner with Shape-Constraint Scene Flow
    • 实例级优化方法
    • 论文：https://cvpr.thecvf.com/virtual/2025/poster/33849
    • 代码：https://github.com/W-QY/SCFlow2
  15. Structure-Aware Correspondence Learning for Relative Pose Estimation
    • 任意物体方法
    • 论文：https://cvpr.thecvf.com/virtual/2025/poster/34892
    • 代码：https://github.com/Cyhhzo02/SAC-Pose-code
  16. UA-Pose: Uncertainty-Aware 6D Object Pose Estimation and Online Object Completion with Partial References
    • 任意物体方法
    • 论文：https://cvpr.thecvf.com/virtual/2025/poster/34276
    • 代码：代码未开源
  17. UNOPose: Unseen Object Pose Estimation with an Unposed RGB-D Reference Image
    • 任意物体方法
    • 论文：https://cvpr.thecvf.com/virtual/2025/poster/33952
    • 代码：https://github.com/shanice-l/UNOPose

NeurIPS

  1. RFMPose: Generative Category-level Object Pose Estimation via Riemannian Flow Matching
    • 类别级方法
    • 论文：https://neurips.cc/virtual/2025/poster/119779
    • 代码：https://github.com/shabiouyang/RMFPose
  2. SingRef6D: Monocular Novel Object Pose Estimation with a Single RGB Reference
    • 任意物体方法
    • 论文：https://neurips.cc/virtual/2025/poster/115633
    • 代码：代码未开源

ICLR

  1. 6D Object Pose Tracking in Internet Videos for Robotic Manipulation
    • 任意物体跟踪方法
    • 论文：https://openreview.net/pdf?id=1CIUkpoata
    • 代码：https://github.com/ponimatkin/freepose
  2. Learning Shape-Independent Transformation via Spherical Representations for Category-Level Object Pose Estimation
    • 类别级方法
    • 论文：https://openreview.net/pdf?id=D4xztKoz0Y
    • 代码：代码未开源

ICML

ECCV

ICCV

  1. 6DOPE-GS: Online 6D Object Pose Estimation using Gaussian Splatting
    • 任意物体方法
    • 论文：https://openaccess.thecvf.com/content/ICCV2025/papers/Jin_6DOPE-GS_Online_6D_Object_Pose_Estimation_using_Gaussian_Splatting_ICCV_2025_paper.pdf
    • 代码：代码未开源
  2. BoxDreamer: Dreaming Box Corners for Generalizable Object Pose Estimation
    • 任意物体方法
    • 论文：https://openaccess.thecvf.com/content/ICCV2025/papers/Yu_BoxDreamer_Dreaming_Box_Corners_for_Generalizable_Object_Pose_Estimation_ICCV_2025_paper.pdf
    • 代码：https://github.com/zju3dv/BoxDreamer
  3. CleanPose: Category-Level Object Pose Estimation via Causal Learning and Knowledge Distillation
    • 类别级方法
    • 论文：https://openaccess.thecvf.com/content/ICCV2025/papers/Lin_CleanPose_Category-Level_Object_Pose_Estimation_via_Causal_Learning_and_Knowledge_ICCV_2025_paper.pdf
    • 代码：https://github.com/chrislin0621/CleanPose
  4. Deterministic Object Pose Confidence Region Estimation
    • 实例级方法
    • 论文：https://arxiv.org/pdf/2506.22720
    • 代码：代码未开源
  5. Environment-Agnostic Pose: Generating Environment-independent Object Representations for 6D Pose Estimation
    • 实例级方法
    • 论文：https://openaccess.thecvf.com/content/ICCV2025/papers/Zhang_Environment-Agnostic_Pose_Generating_Environment-independent_Object_Representations_for_6D_Pose_Estimation_ICCV_2025_paper.pdf
    • 代码：代码未开源
  6. HccePose(BF): Predicting Front & Back Surfaces to Construct Ultra-Dense 2D-3D Correspondences for Pose Estimation
    • 实例级方法
    • 论文：https://openaccess.thecvf.com/content/ICCV2025/papers/Wang_HccePoseBF_Predicting_Front__Back_Surfaces_to_Construct_Ultra-Dense_2D-3D_ICCV_2025_paper.pdf
    • 代码：https://github.com/WangYuLin-SEU/HCCEPose
  7. Joint Learning of Pose Regression and Denoising Diffusion with Score Scaling Sampling for Category-level 6D Pose Estimation
    • 类别级方法
    • 论文：https://openaccess.thecvf.com/content/ICCV2025/papers/Lee_Joint_Learning_of_Pose_Regression_and_Denoising_Diffusion_with_Score_ICCV_2025_paper.pdf
    • 代码：代码未开源
  8. MixRI: Mixing Features of Reference Images for Novel Object Pose Estimation
    • 任意物体方法
    • 论文：https://openaccess.thecvf.com/content/ICCV2025/papers/Liu_MixRI_Mixing_Features_of_Reference_Images_for_Novel_Object_Pose_ICCV_2025_paper.pdf
    • 代码：代码未开源
  9. RayPose: Ray Bundling Diffusion for Template Views in Unseen 6D Object Pose Estimation
    • 任意物体方法
    • 论文：https://openaccess.thecvf.com/content/ICCV2025/papers/Huang_RayPose_Ray_Bundling_Diffusion_for_Template_Views_in_Unseen_6D_ICCV_2025_paper.pdf
    • 代码：代码未开源
  10. Unified Category-Level Object Detection and Pose Estimation from RGB Images using 3D Prototypes
    • 类别级方法
    • 论文：https://openaccess.thecvf.com/content/ICCV2025/papers/Fischer_Unified_Category-Level_Object_Detection_and_Pose_Estimation_from_RGB_Images_ICCV_2025_paper.pdf
    • 代码：https://github.com/Fischer-Tom/unified-detection-and-pose-estimation

AAAI

  1. KeyPose: Category-Level 6D Object Pose Estimation with Self-Adaptive Keypoints
    • 类别级方法
    • 论文：https://ojs.aaai.org/index.php/AAAI/article/view/33046
    • 代码：代码未开源
  2. Universal Features Guided Zero-Shot Category-Level Object Pose Estimation
    • 类别级方法
    • 论文：https://ojs.aaai.org/index.php/AAAI/article/view/32713
    • 代码：代码未开源

IJCAI

  1. Pre-defined Keypoints Promote Category-level Articulation Pose Estimation via Multi-Modal Alignment
    • 类别级铰接物体方法
    • 论文：https://www.ijcai.org/proceedings/2025/0137.pdf
    • 代码：代码未开源
  2. SeqPose: An End-to-End Framework to Unify Single-frame and Video-based RGB Category-Level Pose Estimation
    • 类别级方法
    • 论文：https://www.ijcai.org/proceedings/2025/0237.pdf
    • 代码：代码未开源

WACV

  1. Generalizable Single-View Object Pose Estimation by Two-Side Generating and Matching
    • 论文：https://openaccess.thecvf.com/content/WACV2025/html/Sun_Generalizable_Single-View_Object_Pose_Estimation_by_Two-Side_Generating_and_Matching_WACV_2025_paper.html
    • 代码：https://github.com/scy639/Gen2SM
  2. Shape-Biased Texture Agnostic Representations for Improved Textureless and Metallic Object Detection and 6D Pose Estimation
    • 论文：https://openaccess.thecvf.com/content/WACV2025/html/Honig_Shape-Biased_Texture_Agnostic_Representations_for_Improved_Textureless_and_Metallic_Object_WACV_2025_paper.html
    • 代码：https://github.com/hoenigpeter/randomized_texturing

ICRA

  1. A Unified End-to-End Network for Category-Level and Instance-Level Object Pose Estimation from RGB Images
    • 实例级&类别级方法
    • 论文：https://ieeexplore.ieee.org/document/11128247
    • 代码：代码未开源
  2. MonoDiff9D: Monocular Category-Level 9D Object Pose Estimation via Diffusion Model
    • 类别级方法
    • 论文：https://ieeexplore.ieee.org/document/11127837
    • 代码：https://github.com/CNJianLiu/MonoDiff9D
  3. Occlusion-Aware 6D Pose Estimation with Depth-Guided Graph Encoding and Cross-Semantic Fusion for Robotic Grasping
    • 实例级方法
    • 论文：https://ieeexplore.ieee.org/document/11128248
    • 代码：代码未开源
  4. You Only Estimate Once: Unified, One-stage, Real-Time Category-Level Articulated Object 6D Pose Estimation for Robotic Grasping
    • 类别级铰接物体方法
    • 论文：https://ieeexplore.ieee.org/document/11128513
    • 代码：代码未开源

SIGGRAPH

ACM MM

  1. CLIP-6D: Empowering CLIP as a Zero-Shot 6D Pose Estimator Through Generalizable Object-Specific Representations
    • 任意物体方法
    • 论文：https://dl.acm.org/doi/10.1145/3746027.3754497
    • 代码：代码未开源

AISTATS

CoRL

  1. One View, Many Worlds: Single-Image to 3D object Meets Generative Domain Randomization for One-Shot 6D Pose Estimation
    • 任意物体方法
    • 论文：https://openreview.net/pdf?id=kto4zVmo4w
    • 代码：https://github.com/gzwsama/OnePoseviaGen
  2. PicoPose: Progressive Pixel-to-Pixel Correspondence Learning for Novel Object Pose Estimation
    • 任意物体方法
    • 论文：https://openreview.net/pdf?id=EXgckdYESp
    • 代码：https://github.com/foollh/PicoPose
  3. UnPose: Uncertainty-Guided Diffusion Priors for Zero-Shot Pose Estimation
    • 任意物体方法
    • 论文：https://openreview.net/pdf?id=L5PJBd8ahD
    • 代码：代码未开源

IROS

  1. 6-DoF Object Tracking with Event-based Optical Flow and Frames
    • 实例级跟踪方法
    • 论文：https://ieeexplore.ieee.org/document/11245865
    • 代码：代码未开源
  2. Distilling 3D distinctive local descriptors for 6D pose estimation
    • 任意物体方法
    • 论文：https://ieeexplore.ieee.org/document/11247352
    • 代码：https://github.com/tev-fbk/dGeDi/
  3. DynamicPose: Real-time and Robust 6D Object Pose Tracking for Fast-Moving Cameras and Objects
    • 实例级跟踪方法
    • 论文：https://ieeexplore.ieee.org/document/11247403
    • 代码：https://github.com/Robotics-STAR-Lab/DynamicPose
  4. Generalizable and Actionable Part Detection and Manipulation with SAM-rectified Segmentation and Iterative Pose Refinement
    • 类别级可操作物体方法
    • 论文：https://ieeexplore.ieee.org/document/11246094
    • 代码：代码未开源
  5. Improving 6D Object Pose Estimation of Metallic Household and Industry Objects
    • 类别级方法
    • 论文：https://ieeexplore.ieee.org/document/11247314
    • 代码：代码未开源
  6. MK-Pose: Category-Level Object Pose Estimation via Multimodal-Based Keypoint Learning
    • 类别级方法
    • 论文：https://ieeexplore.ieee.org/document/11247325
    • 代码：https://github.com/yangyifanYYF/MK-Pose
  7. Occlusion-Aware 6D Pose Estimation with Visual Observation Guided Diffusion Model
    • 类别级方法
    • 论文：https://ieeexplore.ieee.org/document/11247112
    • 代码：代码未开源
  8. OmniPose6D: Towards Short-Term Object Pose Tracking in Dynamic Scenes from Monocular RGB
    • 任意物体跟踪方法
    • 论文：https://ieeexplore.ieee.org/document/11246949
    • 代码：代码未开源
  9. RAG-6DPose: Retrieval-Augmented 6D Pose Estimation via Leveraging CAD as Knowledge Base
    • 实例级方法
    • 论文：https://ieeexplore.ieee.org/document/11246449
    • 代码：https://github.com/SresserS/RAG-6DPose-code
  10. RCGNet: RGB-based Category-Level 6D Object Pose Estimation with Geometric Guidance
    • 类别级方法
    • 论文：https://ieeexplore.ieee.org/document/11247442
    • 代码：代码未开源
  11. RGBTrack: Fast, Robust Depth-Free 6D Pose Estimation and Tracking
    • 实例级跟踪方法
    • 论文：https://ieeexplore.ieee.org/document/11247381
    • 代码：https://github.com/GreatenAnoymous/RGBTrack
  12. Uncertainty-Aware Knowledge Distillation for Compact and Efficient 6DoF Pose Estimation
    • 实例级方法
    • 论文：https://ieeexplore.ieee.org/document/11245982
    • 代码：代码未开源
  13. VAPO: Visibility-Aware Keypoint Localization for Efficient 6DoF Object Pose Estimation
    • 实例级方法
    • 论文：https://ieeexplore.ieee.org/document/11247012
    • 代码：https://github.com/RuyiLian/VAPO

ICPR

RSS

  1. V-HOP: Visuo-Haptic 6D Object Pose Tracking
    • 实例级跟踪方法
    • 论文：https://roboticsconference.org/2025/program/papers/37/
    • 代码：https://github.com/brown-ivl/v-hop

IJCNN

BMVC

3DV

  1. GS-Pose: Generalizable Segmentation-Based 6D Object Pose Estimation With 3D Gaussian Splatting
    • 任意物体方法
    • 论文：https://openreview.net/pdf?id=8xHWWXZ9pq
    • 代码：https://github.com/dingdingcai/GSPose

ICIP

  1. Adapting Foundation Features via Cross-View Contrastive Learning for Unseen Object Pose Estimation
    • 任意物体方法
    • 论文：https://ieeexplore.ieee.org/document/11084358
    • 代码：代码未开源

ACCV

ICME

  1. Diff-Art: Category-level Articulation Pose Estimation via Conditional Diffusion
    • 类别级铰接物体方法
    • 论文：https://ieeexplore.ieee.org/document/11209596
    • 代码：代码未开源
  2. GASEM: Boosting Generalized and Actionable Parts Segmentation and Pose Estimation via Object Motion Perception
    • 类别级铰接物体方法
    • 论文：https://ieeexplore.ieee.org/document/11208907
    • 代码：https://github.com/Zirconium233/GASEM
  3. Multi-mode Bidirectional Feature Fusion and Domain-consistency Refinement for Real-time Monocular 6D Object Pose Estimation
    • 实例级方法
    • 论文：https://ieeexplore.ieee.org/document/11208924
    • 代码：代码未开源

ECAI

ISMAR

SIGGRAPH Asia

ICANN

  1. RDGE-6D: Reverse Direction Geometry Injection for 6D Pose Estimation
    • 实例级方法
    • 论文：https://link.springer.com/chapter/10.1007/978-3-032-04546-1_25
    • 代码：代码未开源

ACML