AI Paper Atlas

Research map / 3D vision / Face and shape

ART: Articulated Reconstruction Transformer

Zizhang Li, Cheng Zhang, Zhengqin Li, Henry Howard-Jenkins, Zhaoyang Lv, Chen Geng, Jiajun Wu, Richard Newcombe, Jakob Engel, Zhao Dong

CVPR 2026 · 3D vision › Face and shape · approach: Articulated and non-rigid shape learning · 7 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 static 3D pre-training on 130k part-decomposed objects (ablation)ART · PartNet-mobility full test setPSNR: 27.495 -> 28.678Improved
add static 3D pre-training (ablation)ART · PartNet-mobility full test setChamfer Distance: 0.013 -> 0.008Improved
use multi-view (4 views) instead of monocular input (ablation)ART · PartNet-mobility full test setPSNR: 25.961 -> 28.678Improved
use canonical rest-state instead of first-frame parameterization (ablation)ART · PartNet-mobility full test setPSNR: 23.587 -> 27.495Improved
use canonical rest-state parameterization (ablation)ART · PartNet-mobility full test setd_cDist: 0.113 -> 0.082Improved
add predefined part ordering (ablation)ART · PartNet-mobility full test setPSNR: 22.588 -> 27.495Improved
add predefined part ordering (ablation)ART · PartNet-mobility full test setd_cDist: 0.208 -> 0.082Improved
add per-part rendering losses (ablation)ART · PartNet-mobility full test setChamfer Distance: 0.022 -> 0.008Improved
add per-part rendering losses (ablation)ART · PartNet-mobility full test setPSNR: 24.465 -> 28.678Improved
add per-part rendering losses (ablation)ART · PartNet-mobility full test setd_cDist: 0.075 -> 0.062Improved
use ART instead of SINGAPOART · StorageFurniture test set (monocular)Chamfer Distance: 0.0059 -> 0.0019Improved
use ART instead of DTA (with depth)ART · PartNet-mobility test set (10 sequences, 4 views x 2 states)Chamfer Distance: 0.008 -> 0.009Hurt

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