文献类型：期刊文献

英文题名：MPM-GS: Optimizing Sparse-View 3D Scene Reconstruction with Virtual View Rendering and Multimodal Regularization

作者：Yao, Mingyuan[1,2,3]; Zhang, Shulong[1,2,3]; Huo, Yukang[1,2,3]; Zhao, Jiayin[1,2,3]; Liu, Xiao[1,2,3]; Xue, Lin[4]; Chen, Yingyi[1,2,3]; Wang, Haihua[1,2,3]

第一作者：Yao, Mingyuan

机构：[1] National Innovation Center for Digital Fishery, China; [2] Key Laboratory of Smart Farming Technologies for Aquatic Animal and Livestock, Ministry of Agriculture and Rural Affairs, China; [3] College of Information and Electrical Engineering, China Agricultural University, Beijing, China; [4] Smart City College, Beijing Union University, China

第一机构：National Innovation Center for Digital Fishery, China

年份：2025

外文期刊名：Proceedings of the International Joint Conference on Neural Networks

收录：EI(收录号：20255019674659)

语种：英文

摘要：3D Gaussian Splatting is widely used in 3D reconstruction and has applications in novel view synthesis and scene generation. Recent work has addressed this problem by leveraging multi-view data for high-quality 3D scene reconstruction. Unfortunately, these approaches suffer from overfitting with sparse-view data due to insufficient view information, resulting in artifacts and aliasing. In contrast, we propose a novel method, MPM-GS, which incorporates monocular depth estimation, virtual views, and regularization techniques to address these challenges and improve reconstruction quality under sparse-view conditions. This fixes the overfitting and artifact issues, however, it does not solve all the limitations of sparse-view data in extreme edge cases. Consequently, we develop a novel densification strategy to optimize Gaussian point distributions and improve scene accuracy. While promising, this densification process is non-trivial, as it requires balancing efficiency with rendering fidelity. Therefore, we further optimize Gaussian points and generate new, representative points to enhance both accuracy and computational efficiency. We evaluate MPM-GS both qualitatively and quantitatively on the Tanks & Temples and LLFF datasets, achieving excellent rendering quality and fast rendering speed, particularly in novel view synthesis. ? 2025 IEEE.