AI Paper Atlas

Research map / 3D vision / 3D generation

3R-Adapter: Retrieval, Rewiring, and Refinement for Efficient Adaptation of 3D Reconstruction Model

Zixiang Ai, Zhenyu Cui, Xu Zou, Jiahuan Zhou

NeurIPS 2026 · 3D vision › 3D generation · approach: Diffusion-based novel view synthesis · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Proposes 3R-Adapter, an adaptation framework (retrieval, rewiring, refinement) for feed-forward 3D reconstruction models to handle long-tail scenarios by addressing relational attention structure collapse.

Problem

Feed-forward 3D reconstruction models generalize well under large-scale pre-training but fail in long-tail scenarios such as low-light reconstruction, joint human-scene reconstruction, and test-time adaptation, attributed to collapse of cross-view attention correspondences.

Approach

Identifies relational attention structure collapse (failure of cross-view correspondences) as a primary cause and proposes an efficient adapter built around retrieval, rewiring, and refinement to restore that structure during adaptation.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add Relational Rewiring on top of Deformable Retrieval (ablation)3R-Adapter (CUT3R) · Oxford Day&NightChamfer: 1.152 -> 0.842Improved
add Deformable Retrieval alone (ablation)CUT3R · Oxford Day&NightChamfer: 1.201 -> 1.152Improved
add Consistency Refinement alone (ablation)CUT3R · Oxford Day&NightChamfer: 1.201 -> 1.012Improved
use all three components together (ablation)CUT3R · Oxford Day&NightChamfer: 1.012 -> 0.581Improved
increase sampling locations K from 1 to 4 (ablation)3R-Adapter · Oxford Day&NightChamferImproved
use more than 4 sampling locations (ablation)3R-Adapter · Oxford Day&NightChamferNo clear effect
increase relational fusion factor alpha beyond 0.10 (ablation)3R-Adapter · Oxford Day&NightChamferHurt
overly large search radius rho (ablation)3R-Adapter · Oxford Day&NightChamferHurt
replace channel-wise PEFT with 3R-AdapterCUT3R · Oxford Day&NightChamfer: 0.791 -> 0.581Improved
adapt with 3R-Adapterpi3 night Match@1 · Oxford Day&Night (block 31)Match@1: 52.94 -> 52.16Mixed
adapt with 3R-Adapterpi3 night Match@1 · Oxford Day&Night (block 21)Match@1: 14.38 -> 19.21Improved
use 3R-Adapter vs specialized TTA methodsDUSt3R · 7-Scenes / NRGBDChamfer: 0.241 -> 0.216 (7-Scenes); 0.162 -> 0.205 (NRGBD)Mixed

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 4 of 5 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