AI Paper Atlas

Research map / 3D vision / Human pose estimation

CheckerPose: Progressive Dense Keypoint Localization for Object Pose Estimation with Graph Neural Network

Ruyi Lian, Haibin Ling

ICCV 2023 · 3D vision › Human pose estimation · approach: Learned camera pose estimation · 32 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add GNN layers for keypoint interactions (ablation)CheckerPose · LMADD(-S) 0.05d: 77.8 -> 84.5Improved
predict binary codes progressively with image feature fusion (ablation)CheckerPose · LMADD(-S) 0.05d: 56.9 -> 84.5Improved
add full segmentation mask auxiliary task (ablation)CheckerPose · LMADD(-S) 0.02d: 30.2 -> 35.7Improved
add visible mask auxiliary task (ablation)CheckerPose · LMADD(-S) 0.02d: 34.1 -> 35.7Improved
use Progressive-X instead of RANSAC/PnP (ablation)CheckerPose · LMADD(-S) 0.02d: 31.1 -> 35.7Improved
use HRNet instead of ResNet34 backbone (ablation)CheckerPose · LMADD(-S) 0.1d: 95.6 -> 97.1Improved
filter correspondences outside visible mask for self-occluded textureless objects (ablation)CheckerPose · LM-Omean ADD(-S) recall: 77.1 -> 77.5Improved
filter correspondences outside visible mask for pudding_box (ablation)CheckerPose · YCB-V pudding_boxADD(-S): 66.4 -> 86.5Improved
mask filtering for foam_brick after 380k steps (ablation)CheckerPose · YCB-V foam_brickADD(-S): 87.2 -> 87.2No clear effect
increase number of keypoints N and neighbors k (ablation)CheckerPose · LMADD(-S) 0.05d: N=512,k=20: 84.5; N=256,k=10: 84.2; N=128,k=20: 82.0Mixed
filter with full mask instead of no filter (ablation)CheckerPose · LM-O (can, cat, driller, eggbox)ADD(-S): can 95.2 -> 95.1; cat 62.0 -> 61.3No clear effect

Later papers that compared with this method

2 later accepted papers report CheckerPose as a baseline in their results tables; 1 of them report a better number on most of the results they share with it.

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 4 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy