AI Paper Atlas

Research map / 3D vision / 3D generation

Nano3D: A Training-Free Approach for Efficient 3D Editing Without Masks

Junliang Ye, Shenghao Xie, Ruowen Zhao, Zhengyi Wang, Hongyu Yan, Wenqiang Zu, Lei Ma, Jun Zhu

ICLR 2026 · 3D vision › 3D generation · approach: Feed-forward 3D asset generation and editing · 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 Voxel-Merge on top of FlowEdit (ablation)Nano3D · 3D editing (qualitative)geometry consistencyImproved
add Slat-Merge after Voxel-Merge (ablation)Nano3D · 3D editing (qualitative)appearance consistencyImproved
lower component-size threshold tau from 100 to 50 or 30 (ablation)Voxel-Merge · wing removal maskmask alignment with edited region: tau=100 vs tau=50, tau=30Hurt
use Nano3D instead of TRELLIS repaint-based editing3D editing · 3D editing benchmark (100 cases)FID: 49.57 -> 27.85Improved
use Nano3D-Edit-100k instead of 3D-Alpacaedit dataset · Edit dataset quality (500 sampled pairs)CLIPScore: 28.42 -> 39.71Improved
use inconsistent pre/post-edit 2D images (3D-Alpaca data) (ablation)Nano3D FlowEdit · 3D-Alpaca casesedit reliabilityHurt
apply FlowEdit to Hunyuan2.1 vecset instead of TRELLIS voxels (ablation)FlowEdit · 3D representation choiceediting stabilityHurt

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