PulseAugur
实时 13:50:35
English(EN) Edge Prediction for Roof Wireframe Reconstruction with Transformers

Transformer重建3D屋顶线框,赢得S23DR挑战赛

研究人员开发了一种新颖的基于Transformer的方法,用于从稀疏点云重建3D屋顶线框。该方法受DETR启发,动态地对输入数据进行子采样,并将其与语义和格式塔特征融合。该系统在“HoHo 22k”数据集上取得了0.6476的混合结构得分,在S23DR Challenge 2026中获得第二名。 AI

影响 引入了一种新颖的用于3D重建的Transformer架构,可能提高计算机视觉中的场景理解能力。

排序理由 该集群包含一篇学术论文,详细介绍了新方法及其在特定数据集和挑战上的性能。

在 arXiv cs.CV 阅读 →

AI 生成摘要 · Google Gemini · 来自 2 个来源。 我们如何撰写摘要 →

报道来源 [2]

  1. arXiv cs.CV TIER_1 English(EN) · Gustav Hanning, Ludvig Dill\'en, Jonathan Astermark, Johanna Lidholm, Viktor Larsson ·

    Edge Prediction for Roof Wireframe Reconstruction with Transformers

    arXiv:2606.02406v1 Announce Type: new Abstract: This paper presents a competitive solution to the S23DR Challenge 2026, which aims to reconstruct 3D house roof wireframe models from sparse SfM point clouds and ground-level semantic segmentations and depth maps. Our proposed metho…

  2. arXiv cs.CV TIER_1 English(EN) · Viktor Larsson ·

    Edge Prediction for Roof Wireframe Reconstruction with Transformers

    This paper presents a competitive solution to the S23DR Challenge 2026, which aims to reconstruct 3D house roof wireframe models from sparse SfM point clouds and ground-level semantic segmentations and depth maps. Our proposed method utilizes an end-to-end Transformer encoder-dec…