研究人员提出了两种从全景图像进行三维室内场景重建的新方法。Argus 在新的 Realsee3D 数据集上进行训练，利用学习到的共视模块精确锚定度量世界坐标系，并在姿态、深度和点云重建方面取得了最先进的性能。FastPano3D 提供了一个快速、端到端的框架，可在几秒钟内从单个全景图像生成三维高斯表示，在推理速度和参数效率方面优于以前的方法，同时保持可比的渲染质量。 AI
影响 这些进展有望加速实时三维场景理解和虚拟环境创建的发展。
排序理由 两篇在 arXiv 上发表的学术论文，详细介绍了三维重建的新方法。
AI 生成摘要 · Google Gemini · 来自 4 个来源。 我们如何撰写摘要 →
arXiv:2606.30047v1 Announce Type: new Abstract: Metric feed-forward 3D reconstruction for panoramic data remains under-explored due to the lack of large-scale panoramic RGB-D training data. We present Realsee3D, a hybrid dataset of 10K indoor scenes (1K real, 9K synthetic) with 2…
arXiv:2606.30352v1 Announce Type: new Abstract: Recent advances in 3D scene reconstruction have highlighted the intricate trade-offs among rendering quality, inference efficiency, and data dependency. To address the challenge of rapidly reconstructing detailed 3D indoor scenes fr…
Recent advances in 3D scene reconstruction have highlighted the intricate trade-offs among rendering quality, inference efficiency, and data dependency. To address the challenge of rapidly reconstructing detailed 3D indoor scenes from minimal input, we introduce FastPano3D, an en…
Metric feed-forward 3D reconstruction for panoramic data remains under-explored due to the lack of large-scale panoramic RGB-D training data. We present Realsee3D, a hybrid dataset of 10K indoor scenes (1K real, 9K synthetic) with 299K panoramic viewpoints and precise metric anno…