从视觉数据生成物理资产的3D数字孪生，助力机器人与建造应用。
Digital Twin Generation from Visual Data: A Survey
- 融合3D高斯泼溅、语义分割等技术构建虚拟副本
- 解决遮挡、光照变化等实际场景挑战
- 适合对数字孪生系统开发感兴趣的工程师
本综述探讨了从视觉数据生成数字孪生的最新进展。这些数字孪生——物理资产的虚拟3D复制品——可应用于机器人、媒体内容创作、设计或施工流程。我们分析了多种方法，包括3D高斯泼溅、生成式修复、语义分割和基础模型，强调了各自的优缺点。此外，讨论了遮挡、光照变化和可扩展性等关键挑战，并指出了研究空白、趋势与未来方向。总体而言，本综述旨在全面概述前沿方法及其在现实应用中的影响。
原文摘要 · Abstract (English)
This survey examines recent advances in generating digital twins from visual data. These digital twins - virtual 3D replicas of physical assets - can be applied to robotics, media content creation, design or construction workflows. We analyze a range of approaches, including 3D Gaussian Splatting, generative inpainting, semantic segmentation, and foundation models, highlighting their respective advantages and limitations. In addition, we discuss key challenges such as occlusions, lighting variations, and scalability, as well as identify gaps, trends, and directions for future research. Overall, this survey aims to provide a comprehensive overview of state-of-the-art methodologies and their implications for real-world applications. Awesome Digital Twin: https://awesomedigitaltwin.github.io
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。