AI Paper Atlas

Research map / 3D vision / Human pose estimation

Generative Category-level Object Pose Estimation via Diffusion Models

Jiyao Zhang, Mingdong Wu, Hao Dong

NeurIPS 2023 · 3D vision › Human pose estimation · approach: Category-level object pose estimation · 13 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
rank candidates by energy model and mean-pool instead of random pick (ablation)GenPose · REAL2755deg2cm mAP: 13.5 -> 52.1Improved
add mean pooling over candidates (random ranking) (ablation)GenPose · REAL2755deg2cm mAP: 13.5 -> 49.4Improved
use energy-based ranking instead of random ranking before pooling (ablation)GenPose · REAL2755deg2cm mAP: 49.4 -> 52.1Improved
filter low-energy candidates (delta=60%) instead of keeping all (ablation)GenPose K=50 · REAL27510deg2cm mAP: 69.7 -> 72.4Improved
filter too aggressively (delta=20%) (ablation)GenPose K=50 · REAL27510deg2cm mAP: 72.4 -> 70.2Hurt
increase candidates K from 10 to 50 (ablation)GenPose · REAL27510deg2cm mAP: 67.9 -> 72.4Improved
increase candidates K from 50 to 100 (ablation)GenPose · REAL27510deg2cm mAP: 72.4 -> 72.6No clear effect
generative diffusion formulation instead of regressioncategory-level pose estimation · REAL2755deg5cm mAP: 48.1 -> 60.9 vs RBP-PoseImproved
generative model on unseen symmetric categories vs RBP-PoseGenPose · REAL275 bowl unseen5deg2cm mAP: 0.0 -> 64.5Improved
adapt estimator to tracking by warm-starting from previous frameGenPose · REAL275 tracking5deg5cm / FPS: 71.5 vs 57.2 (5deg5cm); 17.18 vs 89.21 FPSMixed

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 0 of 1 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