CenterSnap | Object Pose Estimation

Object Pose Estimation · 2022 · ICRA · CCF Unranked

CenterSnap: Single-Shot Multi-Object 3D Shape Reconstruction and Categorical 6D Pose and Size Estimation

Represents object instances by spatial centers and jointly predicts 3D shape, 6D pose, and size in one forward pass for CAD-free category-level multi-object scenes.

Paper Code

Research area

Object Pose Estimation

Task level

Category-level

Object setting

Unseen objects

Input modalities

RGB-D

Requirements

camera intrinsicsdepthcategory prior

Full paper metadata Authors, affiliations, research area, datasets, metrics, method tags, and reading status Expand

Authors: Muhammad Zubair Irshad, Thomas Kollar, Michael Laskey, Kevin Stone, Zsolt Kira
Affiliations: Georgia Institute of Technology / Toyota Research Institute
Conference / journal: Conference · ICRA · CCF Unranked
Research area: Object Pose Estimation
Code availability: Code available
Reading status: Overview:Pending Deep Dive:Pending
Task setting: Category-level · Unseen objects
Datasets: ShapeNet NOCS REAL275
Evaluation metrics: 3D IoU n° mcm
Method tags: Dense Correspondences & Coordinate Fields Synthetic Data & Foundation Models Multi-object Pose Estimation Category-level Pose Estimation Absolute Pose Estimation
Requirements: camera intrinsicsdepthcategory prior

Reading Notes

Overview / Deep Dive

Pending src/paper-notes/centersnap/coarse.md

Paper overview notes

Quickly assess the paper's scope, problem setting, core method, experimental findings, and whether to read it in depth.

Reading notes have not yet been added. Markdown notes placed at the path above will appear here once available.

Planned note structure

01 Basic information
02 Positioning in one sentence
03 Method inputs and outputs
04 Problem setting
05 Core motivation
06 Method overview
07 Experimental overview
08 Strengths and limitations
09 Research takeaways

Related papers

2026 · CVPR

3PT

3PT unifies detection, segmentation, and 6-DoF pose with two Transformers for 2D / 3D object-centric understanding. Continuous refinement of object representations and rendered-feature alignment support zero-shot cross-object perception on BOP-H3, BOP-Industrial, and BOP-Classic.

2026 · AAAI

6DAttack

Studies backdoor attacks on PnP-based and end-to-end 6-DoF pose pipelines. Structurally 3D triggers and modified pose labels are injected into training data so that models behave normally on clean inputs but produce controlled incorrect poses when triggered.

2026 · ICRA

Active Tactile Exploration

A violation-implicit loss incorporates physical-constraint violations into joint estimation of freely moving unknown convex rigid-body trajectories and cuboid / 20-vertex polytope shapes. Observed / Fisher information at a local MLE estimates EIG to select the next touch. In simulation, the EIG strategy yields stable errors below 4 mm after 5 cuboid actions; real Trifinger results at step 5 substantially outperform random selection. Convex geometry, local optima, multiple initializations, and tactile / proprioceptive errors limit the method.