CoMoGaussian: Continuous Motion-Aware Gaussian Splatting from Motion-Blurred Images

2025-03-07

AI Summary

To address geometric distortion and inconsistency in 3D Gaussian Splatting (3DGS) reconstructions caused by motion blur, this paper proposes the first Neural Ordinary Differential Equation (Neural ODE)-based method for modeling continuous camera trajectories—departing from conventional discrete-frame assumptions. We further introduce a Continuous Motion Refinement (CMR) transformation that jointly optimizes geometry and radiance fields under physically grounded imaging constraints, enabling robust reconstruction under motion blur. The approach integrates seamlessly into the 3DGS framework and supports real-time rendering. Extensive evaluation across varying blur intensities demonstrates state-of-the-art performance: significant improvements in PSNR and SSIM, up to 32% reduction in LPIPS, more accurate geometric reconstruction, and enhanced generalization capability.

Technology Category

• Computer Vision: Motion & Tracking
• Search and Optimization: Mixed Discrete/Continuous Search
• Machine Learning: Learning with Manifolds

Application Category

• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

3D Gaussian Splatting (3DGS) has gained significant attention for their high-quality novel view rendering, motivating research to address real-world challenges. A critical issue is the camera motion blur caused by movement during exposure, which hinders accurate 3D scene reconstruction. In this study, we propose CoMoGaussian, a Continuous Motion-Aware Gaussian Splatting that reconstructs precise 3D scenes from motion-blurred images while maintaining real-time rendering speed. Considering the complex motion patterns inherent in real-world camera movements, we predict continuous camera trajectories using neural ordinary differential equations (ODEs). To ensure accurate modeling, we employ rigid body transformations, preserving the shape and size of the object but rely on the discrete integration of sampled frames. To better approximate the continuous nature of motion blur, we introduce a continuous motion refinement (CMR) transformation that refines rigid transformations by incorporating additional learnable parameters. By revisiting fundamental camera theory and leveraging advanced neural ODE techniques, we achieve precise modeling of continuous camera trajectories, leading to improved reconstruction accuracy. Extensive experiments demonstrate state-of-the-art performance both quantitatively and qualitatively on benchmark datasets, which include a wide range of motion blur scenarios, from moderate to extreme blur.

Problem

• Reconstructs 3D scenes from motion-blurred images.
• Predicts continuous camera trajectories using neural ODEs.
• Improves reconstruction accuracy with continuous motion refinement.

Innovation

• Continuous Motion-Aware Gaussian Splatting for 3D reconstruction
• Neural ODEs predict continuous camera trajectories
• Continuous Motion Refinement improves motion blur modeling