AI Paper Atlas

Research map / 3D vision / Human pose estimation

Geometry-Preserving Image Generation for 6D Object Pose Estimation

Jiafeng Zhang, Rui Song, Zhengtai Zhang, Jiaojiao Li, Kailang Cao, Lizhang Peng, David Ferstl, Yinlin Hu

ECCV 2026 · 3D vision › Human pose estimation · approach: Instance-level 6D object pose · 0 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 ControlNetGeom geometry conditioning (ablation)GenerationPose · BOP macro avgNCD (%): 1.318 -> 0.867Improved
add ControlNetGeom geometry conditioning (ablation)GenerationPose · BOP macro avgBF1@2: 0.881 -> 0.898Improved
add mask-aware attention (ablation)GenerationPose · BOP macro avgBF1@2: 0.892 -> 0.898Improved
add mask-aware attention (ablation)GenerationPose · BOP macro avgNCD (%): 0.873 -> 0.867Improved
add mask dilation (r>0) at inference (ablation)GenerationPose · BOP macro avgNCD (%): 0.886 -> 0.867Improved
use DDIM strength 0.6 instead of 0.3 (ablation)GenerationPose · BOP macro avgNCD (%): 1.085 -> 0.867Improved
use DDIM strength 0.6 instead of 1.0 (ablation)GenerationPose · BOP macro avgNCD (%): 1.045 -> 0.867Improved
translate renders into pseudo-real imagesGenerationPose · BOP macro avg (FID)FID: 132.84 -> 114.57Improved
translate renders into pseudo-real imagesGenerationPose · BOP macro avg (KID)KID: 0.0986 -> 0.0794Improved
train WDR on generated data instead of PBRWDR · BOP LineMODADI.10d: 40.96 -> 79.87Improved
add generated data to real training dataWDR · BOP LineMODADI.10d: 83.01 -> 85.55Improved
add generated data rather than PBR to real dataWDR · BOP LineMODADI.10d: 83.54 -> 85.55Improved

Earlier methods this paper compares with

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