无需相机参数，通过几何引导生成更对齐的3D/4D场景。
SteerX: Creating Any Camera-Free 3D and 4D Scenes with Geometric Steering
- 在生成阶段统一优化重建，用无姿态模型设计几何奖励函数。
- 零样本推理时提升场景几何对齐度，显著改善生成质量。
- 适合需要高几何一致性但无相机数据的3D内容生成场景。
近期3D/4D场景生成研究强调了视频生成与场景重建中物理对齐的重要性。然而，现有方法在各阶段分别优化对齐，难以处理由其他阶段引发的微小错位问题。本文提出SteerX，一种零样本推理时的几何引导方法，将场景重建统一融入生成过程，通过无姿态前馈重建模型设计两个几何奖励函数，推动数据分布向更好几何对齐方向倾斜。大量实验表明，SteerX能有效提升3D/4D场景生成的几何一致性与整体质量。
原文摘要 · Abstract (English)
Recent progress in 3D/4D scene generation emphasizes the importance of physical alignment throughout video generation and scene reconstruction. However, existing methods improve the alignment separately at each stage, making it difficult to manage subtle misalignments arising from another stage. Here, we present SteerX, a zero-shot inference-time steering method that unifies scene reconstruction into the generation process, tilting data distributions toward better geometric alignment. To this end, we introduce two geometric reward functions for 3D/4D scene generation by using pose-free feed-forward scene reconstruction models. Through extensive experiments, we demonstrate the effectiveness of SteerX in improving 3D/4D scene generation.
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。