TWINGS: Thin Plate Splines Warp-aligned Initialization for Sparse-View Gaussian Splatting

2026-05-21

AI Summary

This work addresses the degradation in reconstruction quality for novel view synthesis under sparse input views, which stems from the sparsity of reconstructed point clouds. To mitigate this issue, the authors propose a non-rigid alignment method based on Thin Plate Splines (TPS), introducing it for the first time into the initialization stage of 3D Gaussian Splatting. By applying a globally consistent TPS deformation between back-projected points and triangulated 3D control points, the method generates a geometrically accurate initial set of Gaussians that preserves color fidelity while significantly enhancing structural detail. Experimental results demonstrate that the proposed approach consistently outperforms existing methods across the DTU, LLFF, and Mip-NeRF360 datasets, achieving high-quality 3D reconstructions from sparse viewpoints.

Abstract

Novel view synthesis from sparse-view inputs poses a significant challenge in 3D computer vision, particularly for achieving high-quality scene reconstructions with limited viewpoints. We introduce TWINGS, a framework that enhances 3D Gaussian Splatting (3DGS) by directly addressing point sparsity. We employ Thin Plate Splines (TPS), a smooth non-rigid deformation model that minimizes bending energy to estimate a globally coherent warp from control-point correspondences, to align backprojected points from estimated depth with triangulated 3D control points, yielding calibrated backprojected points. By sampling these calibrated points near the control points, TWINGS provides a fast and geometrically accurate initialization for 3DGS, ultimately improving structural detail preservation and color fidelity in reconstructed scenes. Extensive experiments on DTU, LLFF, and Mip-NeRF360 demonstrate that TWINGS consistently outperforms existing methods, delivering detailed and accurate reconstructions under sparse-view scenarios.

Problem

• novel view synthesis
• sparse-view
• 3D reconstruction
• scene reconstruction
• 3D computer vision

Innovation

• Thin Plate Splines
• Sparse-View Reconstruction
• 3D Gaussian Splatting
• Non-rigid Registration
• Novel View Synthesis