6D Pose Estimation Using Rgb On Occlusion

Métriques

Mean ADD

Résultats

Résultats de performance de divers modèles sur ce benchmark

		Paper Title
SO-Pose	62.3	SO-Pose: Exploiting Self-Occlusion for Direct 6D Pose Estimation
RNNPose (Trained with synthetic data and LINEMOD training set, w/o pbr data)	60.65	RNNPose: Recurrent 6-DoF Object Pose Refinement with Robust Correspondence Field Estimation and Pose Optimization
GDR-Net	56.1	GDRNPP: A Geometry-guided and Fully Learning-based Object Pose Estimator
DeepIM (Train on Occlusion LineMOD)	55.5	DeepIM: Deep Iterative Matching for 6D Pose Estimation
PPC (Refined from initial PVNet pose)	55.33	Pose Proposal Critic: Robust Pose Refinement by Learning Reprojection Errors
RePOSE	51.6	RePOSE: Fast 6D Object Pose Refinement via Deep Texture Rendering
HybridPose	47.5	HybridPose: 6D Object Pose Estimation under Hybrid Representations
E2E6DoF	47.4	End-to-End Differentiable 6DoF Object Pose Estimation with Local and Global Constraints
DPOD	47.25	DPOD: 6D Pose Object Detector and Refiner
ROPE	45.95	Occlusion-Robust Object Pose Estimation with Holistic Representation
PVNet	40.77	PVNet: Pixel-wise Voting Network for 6DoF Pose Estimation
SegDriven	27	Segmentation-driven 6D Object Pose Estimation
CullNet	24.48	CullNet: Calibrated and Pose Aware Confidence Scores for Object Pose Estimation

0 of 13 row(s) selected.

6D Pose Estimation Using Rgb On Occlusion | SOTA | HyperAI