PrITTI: Primitive-based Generation of Controllable and Editable 3D Semantic Urban Scenes 文章

ArXiv CS.CV2026-05-26NEWSen作者: Christina Ourania Tze, Daniel Dauner, Yiyi Liao, Dzmitry Tsishkou, Andreas Geiger
查看原文 →
关系图谱

PrITTI: Primitive-based Generation of Controllable and Editable 3D Semantic Urban Scenes · 相关技术

相关技术

diffusionginORMODEPLA扩散模型divide-and-conquer partitioningVerseUCTStraight-Through EstimatorSpatial Pivot-Aligned Coordinate-free Embedding (SPACE)SPASEMParts-of-Speech (POS) tagsPPRMITHISForEffort Metric AttentionDiTDenseANNTable