DKPMV: Dense Keypoints Fusion from Multi-View RGB Frames for 6D Pose Estimation of Textureless Objects

2025/10/13 by Chen, Jiahong, Wang, Jinghao, Wang, Zi +3
#Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #Robotics (cs.RO)

paper · doi:10.48550/arxiv.2510.10933

Abstract

6D pose estimation of textureless objects is valuable for industrial robotic applications, yet remains challenging due to the frequent loss of depth information. Current multi-view methods either rely on depth data or insufficiently exploit multi-view geometric cues, limiting their performance. In this paper, we propose DKPMV, a pipeline that achieves dense keypoint-level fusion using only multi-view RGB images as input. We design a three-stage progressive pose optimization strategy that leverages dense multi-view keypoint geometry information. To enable effective dense keypoint fusion, we enhance the keypoint network with attentional aggregation and symmetry-aware training, improving prediction accuracy and resolving ambiguities on symmetric objects. Extensive experiments on the ROBI dataset demonstrate that DKPMV outperforms state-of-the-art multi-view RGB approaches and even surpasses the RGB-D methods in the majority of cases. The code will be available soon.

Citations

• XYZ-IBD: A High-precision Bin-picking Dataset for Object 6D Pose Estimation Capturing Real-world Industrial Complexity
• Reasoning and Learning a Perceptual Metric for Self-Training of Reflective Objects in Bin-Picking with a Low-cost Camera
• Active 6D Pose Estimation for Textureless Objects using Multi-View RGB Frames
• ZeroBP: Learning Position-Aware Correspondence for Zero-shot 6D Pose Estimation in Bin-Picking
• Deep Learning-Based Object Pose Estimation: A Comprehensive Survey
• ASGrasp: Generalizable Transparent Object Reconstruction and 6-DoF Grasp Detection from RGB-D Active Stereo Camera
• 6D-Diff: A Keypoint Diffusion Framework for 6D Object Pose Estimation
• TEMPO: Efficient Multi-View Pose Estimation, Tracking, and Forecasting
• Multi-View Keypoints for Reliable 6D Object Pose Estimation
• CheckerPose: Progressive Dense Keypoint Localization for Object Pose Estimation with Graph Neural Network
• EPro-PnP: Generalized End-to-End Probabilistic Perspective-n-Points for Monocular Object Pose Estimation
• 6D Pose Estimation for Textureless Objects on RGB Frames using Multi-View Optimization
• TransNet: Category-Level Transparent Object Pose Estimation
• Symmetry and Uncertainty-Aware Object SLAM for 6DoF Object Pose Estimation
• ZebraPose: Coarse to Fine Surface Encoding for 6DoF Object Pose Estimation
• Iterative Corresponding Geometry: Fusing Region and Depth for Highly Efficient 3D Tracking of Textureless Objects
• OVE6D: Object Viewpoint Encoding for Depth-based 6D Object Pose Estimation
• Next-Best-View Prediction for Active Stereo Cameras and Highly Reflective Objects
• KDFNet: Learning Keypoint Distance Field for 6D Object Pose Estimation
• SO-Pose: Exploiting Self-Occlusion for Direct 6D Pose Estimation
• A Multi-Hypothesis Approach to Pose Ambiguity in Object-Based SLAM
• ROBI: A Multi-View Dataset for Reflective Objects in Robotic Bin-Picking
• CloudAAE: Learning 6D Object Pose Regression with On-line Data Synthesis on Point Clouds
• Vid2CAD: CAD Model Alignment using Multi-View Constraints from Videos
• CosyPose: Consistent multi-view multi-object 6D pose estimation
• Single Shot 6D Object Pose Estimation
• 6D Object Pose Regression via Supervised Learning on Point Clouds
• KeyPose: Multi-View 3D Labeling and Keypoint Estimation for Transparent Objects
• Single-Stage 6D Object Pose Estimation
• PVN3D: A Deep Point-wise 3D Keypoints Voting Network for 6DoF Pose Estimation
• On Object Symmetries and 6D Pose Estimation from Images
• PoseRBPF: A Rao-Blackwellized Particle Filter for 6D Object Pose Tracking
• DenseFusion: 6D Object Pose Estimation by Iterative Dense Fusion
• Implicit 3D Orientation Learning for 6D Object Detection from RGB Images
• Dynamic Graph CNN for Learning on Point Clouds
• Dynamic Graph CNN for Learning on Point Clouds
• 6-DoF Object Pose from Semantic Keypoints
• Random sample consensus

Related

• KGpose: Keypoint-Graph Driven End-to-End Multi-Object 6D Pose Estimation via Point-Wise Pose Voting
• Active 6D Pose Estimation for Textureless Objects using Multi-View RGB Frames
• 6D Pose Estimation for Textureless Objects on RGB Frames using Multi-View Optimization
• DLTPose: 6DoF Pose Estimation From Accurate Dense Surface Point Estimates
• Multi-View Keypoints for Reliable 6D Object Pose Estimation