Dental3R: Geometry-Aware Pairing for Intraoral 3D Reconstruction from Sparse-View Photographs

2025-11-18

AI Summary

In remote orthodontics, high-fidelity 3D reconstruction from sparse intraoral smartphone images (standard three-view setup) remains challenging due to severe pose-geometry coupling induced by large baseline disparities, inconsistent illumination, and specular reflections. To address this, we propose a graph-guided 3D Gaussian Splatting reconstruction framework that requires no initial pose estimation. We introduce a geometry-aware image-pair selection strategy to enhance robustness in pose estimation and point cloud initialization under sparse-view conditions. Additionally, we incorporate discrete wavelet transform regularization to suppress low-frequency bias caused by photometric supervision, explicitly preserving high-frequency details—such as enamel boundaries and interproximal regions. Evaluated on 950 clinical cases and a 195-video test set, our method achieves significantly superior novel-view synthesis quality and occlusion visualization accuracy compared to state-of-the-art approaches.

Technology Category

• Computer Vision: 3D Computer Vision
• Machine Learning: Feature Construction/Reformulation
• Intelligent Robots: State Estimation

Application Category

• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

Intraoral 3D reconstruction is fundamental to digital orthodontics, yet conventional methods like intraoral scanning are inaccessible for remote tele-orthodontics, which typically relies on sparse smartphone imagery. While 3D Gaussian Splatting (3DGS) shows promise for novel view synthesis, its application to the standard clinical triad of unposed anterior and bilateral buccal photographs is challenging. The large view baselines, inconsistent illumination, and specular surfaces common in intraoral settings can destabilize simultaneous pose and geometry estimation. Furthermore, sparse-view photometric supervision often induces a frequency bias, leading to over-smoothed reconstructions that lose critical diagnostic details. To address these limitations, we propose extbf{Dental3R}, a pose-free, graph-guided pipeline for robust, high-fidelity reconstruction from sparse intraoral photographs. Our method first constructs a Geometry-Aware Pairing Strategy (GAPS) to intelligently select a compact subgraph of high-value image pairs. The GAPS focuses on correspondence matching, thereby improving the stability of the geometry initialization and reducing memory usage. Building on the recovered poses and point cloud, we train the 3DGS model with a wavelet-regularized objective. By enforcing band-limited fidelity using a discrete wavelet transform, our approach preserves fine enamel boundaries and interproximal edges while suppressing high-frequency artifacts. We validate our approach on a large-scale dataset of 950 clinical cases and an additional video-based test set of 195 cases. Experimental results demonstrate that Dental3R effectively handles sparse, unposed inputs and achieves superior novel view synthesis quality for dental occlusion visualization, outperforming state-of-the-art methods.

Problem

• Reconstructing 3D dental models from sparse smartphone photographs for remote orthodontics
• Addressing unstable geometry estimation from large view baselines and specular surfaces
• Preventing over-smoothed reconstructions that lose critical diagnostic dental details

Innovation

• Geometry-Aware Pairing Strategy for image pairs
• Wavelet-regularized objective with 3D Gaussian Splatting
• Pose-free graph-guided pipeline for sparse photographs