AI Paper Atlas

Research map / 3D vision / 3D generation

DreamPartGen: Semantically Grounded Part-Level 3D Generation via Collaborative Latent Denoising

Tianjiao (Joey) Yu, Xinzhuo Li, Muntasir Wahed, Jerry Xiong, Yifan Shen, Ying Shen, Ismini Lourentzou

ECCV 2026 · 3D vision › 3D generation · approach: Text-to-3D scene generation · 3 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
remove global relational tokens S_glb and object-level sync (add them) (ablation)DreamPartGen · PartRel3D subsetCD: 0.292 -> 0.145Improved
add local semantic tokens S_loc co-denoised with part latents (ablation)DreamPartGen · PartRel3D subsetCD: 0.781 -> 0.145Improved
add learnable part identifier embedding (ablation)DreamPartGen · PartRel3D subsetIoU: 0.438 -> 0.212Improved
add persistent RSLs for text-shape alignment (ablation)DreamPartGen · PartRel3D subsetULIP-T: 0.084 -> 0.158Improved
add spatial triplets to text conditioning (ablation)DreamPartGen · PartRel3DCD: 0.167 -> 0.112Improved
add functional triplets alone instead of text (ablation)DreamPartGen · PartRel3DCD: 0.167 -> 0.180Hurt
add image to Text+FT+ST (ablation)DreamPartGen · PartRel3DCD: 0.085 -> 0.080Improved
increase local RSL tokens beyond 16 (ablation)DreamPartGen · PartRel3DCD: 0.084 -> 0.085No clear effect
use VLM-parsed relations at inference instead of prompt-only (ablation)DreamPartGen · PartRel3DCD: 0.099 vs 0.101 (Qwen2.5-VL)No clear effect
test on held-out relation predicatesDreamPartGen · PartRel3D OOD-relr-FID: 9.783 -> 10.631Mixed
DreamPartGen vs PartCrafter on ShapeNet part independenceDreamPartGen · ShapeNetIoU: 0.499 vs 0.503No clear effect
part-aware co-denoising vs baselines on text-shape alignmentDreamPartGen · PartVerse object-levelULIP-T: 0.164 -> 0.197Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 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