6DoF / model-point distance · error

ADD-S

Average Closest Point Distance

Relaxes point identity: each ground-truth point is matched to the nearest point in the predicted model.

ADD-S uses nearest-point matching to reduce ambiguity from rotational symmetry or repeated structures. It is common on datasets such as YCB-Video.

Explore principles and formulas View related papers 75

Principle

What error does this metric penalize?

The formulas define the computation. The diagrams show whether the error concerns a 3D model, visible surfaces, the image plane, or aggregation across thresholds.

Rotating an axially symmetric cup by 180° may yield an equivalent pose; ADD-S no longer requires the same CAD point to align with its exact counterpart.

It can also reduce errors for locally repeated shapes that are not strictly symmetric.

\operatorname{ADD\!\text{-}\!S}=\frac{1}{|\mathcal M|}\sum_{x_1\in\mathcal M}\min_{x_2\in\mathcal M}\left\lVert (R_{gt}x_1+t_{gt})-(\hat R x_2+\hat t)\right\rVert_2

Nearest-point searches are typically accelerated with a KD-tree or GPU neighbor queries. Its tolerance comes from allowing x_1 and x_2 to differ.

Illustration created for this site to explain the metric's error space; it does not show measured data or a figure from the original paper.

How to compute

Computation steps

01
Transform the ground-truth and predicted model point sets separately.
02
For each ground-truth point, find the nearest neighbor in the predicted point set.
03
Average nearest-neighbor distances, then obtain error, recall, or AUC according to the protocol.

Strength

What it is useful for

More appropriate than ADD for common symmetric objects.
Retains an intuitive distance interpretation in 3D model space.

Caveat

Cautions when reading results

Nearest-point matching can conceal texture-defined orientation differences, such as those of a labeled cylinder.
Unlike BOP, it does not use an explicit global symmetry set; check the protocol for complex discrete symmetries.

Paper atlas

Which papers report this metric?

Links are derived by normalizing the metric names used in each paper. Directly reported metrics are distinguished from metrics associated through a broader protocol.

Filter the paper library

75Related papers

59Directly reported

16Linked through a protocol

2026 Directly reported

Aleatoric Uncertainty from AI-based 6D Object Pose Predictors for Object-relative State Estimation

ICRA / IEEE RA-L · Thomas Jantos, Stephan Weiss, Jan Steinbrener

ADD-S

What error does this metric penalize?

Computation steps

What it is useful for

Cautions when reading results

Which papers report this metric?

Aleatoric Uncertainty from AI-based 6D Object Pose Predictors for Object-relative State Estimation

Learning 6D Object Pose Estimation With Event Cameras Using Synthetic Data and Domain Randomization

Color-Pair Guided Robust Zero-Shot 6D Pose Estimation and Tracking of Cluttered Objects on Edge Devices

ConceptPose: Training-Free Zero-Shot Object Pose Estimation using Concept Vectors

CoordAR: One-Reference 6D Pose Estimation of Novel Objects via Autoregressive Coordinate Map Generation

Cov2Pose: Leveraging Spatial Covariance for Direct Manifold-aware 6-DoF Object Pose Estimation

DynOPETs: A Versatile Benchmark for Dynamic Object Pose Estimation and Tracking in Moving Camera Scenarios

EgoXtreme: A Dataset for Robust Object Pose Estimation in Egocentric Views under Extreme Conditions

Event6D: Event-based Novel Object 6D Pose Tracking

GenHOI: Generalized Hand-Object Pose Estimation with Occlusion Awareness

GraspClutter6D: A Large-scale Real-world Dataset for Robust Perception and Grasping in Cluttered Scenes

HIPPo: Harnessing Image-to-3D Priors for Model-Free Zero-Shot 6D Pose Estimation

Towards in-the-wild Egocentric 3D Hand-Object Pose Estimation

HRDexDB: A Paired Human-Robot Dataset for Cross-Embodiment Dexterous Grasping

Deep Learning-Based Object Pose Estimation: A Comprehensive Survey

OrienPose: Orientation-Guided Novel View Synthesis for Single-Image Unseen Object Pose Estimation

Point2Pose: Occlusion-Recovering 6D Pose Tracking and 3D Reconstruction for Multiple Unknown Objects Via 2D Point Trackers

PoseGaussian: 6D Pose Estimation for Unseen Objects via Sparse-View Object-Level 3D Gaussian Splatting

TwinTrack: Bridging Vision and Contact Physics for Real-Time Tracking of Unknown Objects in Contact-Rich Scenes

VPHO: Joint Visual-Physical Cue Learning and Aggregation for Hand-Object Pose Estimation

6DOPE-GS: Online 6D Object Pose Estimation using Gaussian Splatting

Any6D: Model-free 6D Pose Estimation of Novel Objects

BoxDreamer: Dreaming Box Corners for Generalizable Object Pose Estimation

CRISP: Object Pose and Shape Estimation with Test-Time Adaptation

Deterministic Object Pose Confidence Region Estimation

DynamicPose: Real-time and Robust 6D Object Pose Tracking for Fast-Moving Cameras and Objects

Environment-Agnostic Pose: Generating Environment-independent Object Representations for 6D Pose Estimation

GS-Pose: Generalizable Segmentation-Based 6D Object Pose Estimation With 3D Gaussian Splatting

iG-6DoF: Model-free 6DoF Pose Estimation for Unseen Object via Iterative 3D Gaussian Splatting

JENGA: Object Selection and Pose Estimation for Robotic Grasping from a Stack

Visuo-Tactile Object Pose Estimation for a Multi-Finger Robot Hand with Low-Resolution In-Hand Tactile Sensing

Occlusion-Aware 6D Pose Estimation with Depth-Guided Graph Encoding and Cross-Semantic Fusion for Robotic Grasping

ONDA-Pose: Occlusion-Aware Neural Domain Adaptation for Self-Supervised 6D Object Pose Estimation

One2Any: One-Reference 6D Pose Estimation for Any Object

RayPose: Ray Bundling Diffusion for Template Views in Unseen 6D Object Pose Estimation

SCFlow2: Plug-and-Play Object Pose Refiner with Shape-Constraint Scene Flow

UA-Pose: Uncertainty-Aware 6D Object Pose Estimation and Online Object Completion with Partial References

Uncertainty-Aware Knowledge Distillation for Compact and Efficient 6DoF Pose Estimation

VAPO: Visibility-Aware Keypoint Localization for Efficient 6DoF Object Pose Estimation

ViTa-Zero: Zero-Shot Visuotactile Object 6D Pose Estimation

ZeroBP: Learning Position-Aware Correspondence for Zero-Shot 6D Pose Estimation in Bin-Picking

ZeroPose: CAD-Prompted Zero-Shot Object 6D Pose Estimation in Cluttered Scenes

6D-Diff: A Keypoint Diffusion Framework for 6D Object Pose Estimation

Learning to Estimate 6DoF Pose from Limited Data: A Few-Shot, Generalizable Approach using RGB Images

FAFA: Frequency-Aware Flow-Aided Self-Supervision for Underwater Object Pose Estimation

FoundationPose: Unified 6D Pose Estimation and Tracking of Novel Objects

HiPose: Hierarchical Binary Surface Encoding and Correspondence Pruning for RGB-D 6DoF Object Pose Estimation

Learning Better Keypoints for Multi-Object 6DoF Pose Estimation

L4D-Track: Language-to-4D Modeling Towards 6-DoF Tracking and Shape Reconstruction in 3D Point Cloud Stream

Effects of Markers in Training Datasets on the Accuracy of 6D Pose Estimation

MFOS: Model-Free & One-Shot Object Pose Estimation

MOTPose: Multi-Object 6D Pose Estimation for Dynamic Video Sequences Using Attention-Based Temporal Fusion

MRC-Net: 6-DoF Pose Estimation with MultiScale Residual Correlation

NeRF-Feat: 6D Object Pose Estimation using Feature Rendering

NeuSurfEmb: A Complete Pipeline for Dense Correspondence-Based 6D Object Pose Estimation without CAD Models

Open-Vocabulary Object 6D Pose Estimation

PACE: A Large-Scale Dataset with Pose Annotations in Cluttered Environments

Pseudo-keypoint RKHS Learning for Self-supervised 6DoF Pose Estimation

SAM-6D: Segment Anything Model Meets Zero-Shot 6D Object Pose Estimation

Scalable Network and Adaptive Refinement Module for 6D Pose Estimation of Diverse Industrial Components

YOLO-6D-Pose: Enhancing YOLO for Single-Stage Monocular Multi-Object 6D Pose Estimation

MV6D: Multi-View 6D Pose Estimation on RGB-D Frames Using a Deep Point-wise Voting Network

PIZZA: A Powerful Image-only Zero-Shot Zero-CAD Approach to 6 DoF Tracking

YOLOPose: Transformer-based Multi-Object 6D Pose Estimation using Keypoint Regression

FFB6D: A Full Flow Bidirectional Fusion Network for 6D Pose Estimation

GDR-Net: Geometry-Guided Direct Regression Network for Monocular 6D Object Pose Estimation

CosyPose: Consistent Multi-view Multi-object 6D Pose Estimation

PVN3D: A Deep Point-wise 3D Keypoints Voting Network for 6DoF Pose Estimation

DenseFusion: 6D Object Pose Estimation by Iterative Dense Fusion

DPOD: 6D Pose Object Detector and Refiner

PVNet: Pixel-wise Voting Network for 6DoF Pose Estimation

Deep Object Pose Estimation for Semantic Robotic Grasping of Household Objects

PoseCNN: A Convolutional Neural Network for 6D Object Pose Estimation in Cluttered Scenes

Real-Time Seamless Single Shot 6D Object Pose Prediction