FreeView: Robust 3D Reconstruction from Uncurated Views

Junho Kim; Hyunyoung Choi; Seongwon Lee

cs.CVOct 24, 2026

FreeView: Robust 3D Reconstruction from Uncurated Views

Authors: Junho Kim, Hyunyoung Choi, Seongwon Lee

Organizations: School of Electrical Engineering, Kookmin University, Seoul 02707, South Korea

Abstract

Reconstructing a coherent 3D scene from uncurated image collections remains challenging due to the presence of irrelevant or non-overlapping views. Recent work has shown that feed-forward reconstruction models contain internal view-consistency signals that support distractor rejection. However, reference-conditioned filtering relies on a curated clean reference, which is not guaranteed in practical image collections. To address this limitation, we introduce FreeView, a training-free framework that jointly selects a filtering reference and rejects distractor views from an uncurated collection. FreeView leverages collection-level consensus to identify a representative view that serves as a reliable reference for filtering. It then adapts view rejection to each input collection by combining complementary visual and geometric evidence. This approach preserves useful views under substantial appearance and viewpoint variations while rejecting distractors. FreeView integrates with both VGGT and pi^3, removing the need for a manually curated reference or manual calibration of view rejection. Extensive experiments across multiple datasets demonstrate high clean-view retention and distractor rejection rates, leading to reliable 3D reconstruction from uncurated views.

Figures & tables

Fig. 2 : Overview of FreeView. Given an uncurated image collection, Consensus Reference Selection identifies a representative view through agreement among related images, establishing a reference for filtering without prior knowledge of which views are clean. Adaptive Dual-Cue Rejection combines complementary visual and geometric evidence to adapt view selection to each collection. Retaining views supported by either cue preserves useful observations under appearance and viewpoint variations, while rejecting views unsupported by both suppresses distractors. The retained views are then processed by VGGT or $\pi^{3}$ for final reconstruction.

TABLE I : View filtering and pose estimation across distractor levels. For the same mixed image collections, reference-conditioned methods are evaluated with either a clean or a distractor reference to assess sensitivity to reference choice, while FreeView selects its reference automatically. CK/DR denote clean-view retention/distractor rejection; T/R denote translation/rotation recall over all clean views, counting rejected views as failures. For

\pi^{3}

, the reference is used only for filtering. Bold / underline indicate the best/second-best non-oracle results per backbone. † denotes DINOv2 features from the VGGT encoder.

CRS	ADCR	CK $\uparrow$	DR $\uparrow$	T@2/5/10% $\uparrow$	R@2/5/10 ${}^{\circ}\uparrow$
Phototourism
✗	✗	0.1	78.9	0.0 / 0.0 / 0.0	0.0 / 0.0 / 0.0
✓	✗	100.0	99.0	52.8 / 81.7 / 93.3	97.2 / 100.0 / 100.0
✓	✓	100.0	99.7	53.2 / 82.0 / 93.8	97.3 / 100.0 / 100.0
ETH3D
✗	✗	0.7	86.4	0.4 / 0.4 / 0.4	0.0 / 0.0 / 0.0

TABLE II : Ablation study on Consensus Reference Selection (CRS) and Adaptive Dual-Cue Rejection (ADCR).

Fig. 3 : Effect of neighborhood size on reference selection and view filtering on Phototourism under the Large setting, with 30 clean views and 50 distractors.

Filtering Cue		Fusion	CK $\uparrow$	DR $\uparrow$	T@2/5/10% $\uparrow$	R@2/5/10 ${}^{\circ}\uparrow$
DINOv2	VGGT Feat.	Fusion	CK $\uparrow$	DR $\uparrow$	T@2/5/10% $\uparrow$	R@2/5/10 ${}^{\circ}\uparrow$
Phototourism
✓	✗	–	98.9	100.0	52.4 / 81.1 / 94.2	96.2 / 98.9 / 98.9
✗	✓	–	91.7	99.7	47.2 / 75.3 / 85.4	84.4 / 91.7 / 91.7
✓	✓	OR	85.0	100.0	42.9 / 70.0 / 80.3	77.8 / 85.0 / 85.0
✓	✓	Avg	62.8	100.0	40.4 / 56.7 / 62.2	56.1 / 62.8 / 62.8

TABLE III : Comparison of view filtering cues and fusion strategies.

3D Reconstruction

3D reconstruction recovers a scene’s shape, appearance, and camera motion from images, video, depth, or other sensor measurements.

See all 957 papers

Visual Geometry Transformer in the Wild: Distractor-Free 3D Reconstruction
Jun 22, 2026Tianbo Pan, Xingyi Yang, Shizun Wang +13D ViTs 3D Reconstruction
National University of Singapore · The Hong Kong Polytechnic University
Per-View Gaussian Predictions Enable Training-Free Distractor Filtering in Feed-Forward 3DGS
Aug 27, 2026Kangmin Seo, Jae-Pil Heo3D Reconstruction 3D Gaussian Splatting
Sungkyunkwan University
AnyRecon: Arbitrary-View 3D Reconstruction with Video Diffusion Model
Apr 21, 2026Yutian Chen, Shi Guo, Renbiao Jin +7Video Diffusion Models 3D Reconstruction
Shanghai AI Lab · Shanghai Jiao Tong University · The University of Hong Kong +2

Jun 22, 2026J/K move · Enter open · S save

Visual Geometry Transformer in the Wild: Distractor-Free 3D Reconstruction

Tianbo Pan, Xingyi Yang, Shizun Wang, Xinchao Wang

National University of Singapore · The Hong Kong Polytechnic University

3D ViTs 3D Reconstruction Multi-View 3D Reconstruction Feed-Forward 3D Reconstruction

Open paper AskPDF arXiv

Current end-to-end multi-view 3D reconstruction methods achieve impressive results, but rely on a restrictive static assumption: the scenes is entire distractor-free with perfect cross-view geometry. This reliance on idealized inputs causes even the most advanced methods to fail in real-world settings, where transient distractors and occlusions present. To address this, we propose Visual Geometry Transformer in the Wild (VGTW), an end-to-end framework for robust reconstruction from inconsistent views. At its core, we isolate and suppress distractor-affected regions while preserving the consistent components across views. Specifically, we introduce a Distractor-aware Training (DAT) strategy that separates clean features from distractor-contaminated ones in the attention mechanism while enforcing feature consistency across images. To enable this, we train the model with an auxiliary mask prediction head, using supervision from a new dataset we collected with pixel-level distractor masks. The resulting VGTW model is a feed-forward network that directly outputs clean, distractor-free point clouds. Remarkably, it requires no additional 3D supervision, remains computationally efficient, and is compatible with existing pipelines. Extensive experiments validate our approach, demonstrating state-of-the-art performance and robust generalization in diverse, real-world scenarios.

Abstract

3D Reconstruction

Explore similar work

Visual Geometry Transformer in the Wild: Distractor-Free 3D Reconstruction

Per-View Gaussian Predictions Enable Training-Free Distractor Filtering in Feed-Forward 3DGS

AnyRecon: Arbitrary-View 3D Reconstruction with Video Diffusion Model