DGGT：使用无位姿图像对动态驾驶场景进行前馈4D重建
DGGT: Feedforward 4D Reconstruction of Dynamic Driving Scenes using Unposed Images
将相机位姿作为模型输出而非输入，用前馈网络从稀疏无位姿图像直接预测逐帧3D高斯图和相机参数，并引入动态头与寿命头解耦动态、保持时序一致性，为自动驾驶场景的快速4D重建与重仿真提供了可复用的无位姿前馈范式。
原文依据 we reformulate pose as an output of the model, enabling reconstruction directly from sparse, unposed images
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论自动驾驶需要快速、可扩展的4D重建与重仿真来支持训练和评估，然而大多数动态驾驶场景方法仍依赖逐场景优化、已知相机标定或短帧窗口，导致速度慢且不实用。我们从���馈视角重新审视该问题，提出驾驶高斯接地Transformer（DGGT），一个用于无位姿动态场景重建的统一框架。我们注意到现有公式将相机位姿视为必需输入，限制了灵活性和可扩展性。相反，我们将位姿重新表述为模型的输出，从而能够直接从稀疏、无位姿图像进行重建，并支持长序列的任意数量视图。我们的方法联合预测逐帧3D高斯图和相机参数，通过轻量级动态头解耦动态，并利用寿命头调制随时间变化的可见性以保持时序一致性。基于扩散的渲染精炼进一步减少运动/插值伪影，并在稀疏输入下提升新视角质量。最终得到一个单次前馈、无位姿的算法，实现了最优的性能和速度。在大规模驾驶基准（Waymo、nuScenes、Argoverse2）上训练和评估，我们的方法在每个数据集上训练以及跨数据集零样本迁移时均优于先前工作，并且随着输入帧数增加具有良好的扩展性。
英文原文摘要 1475 字符
Autonomous driving needs fast, scalable 4D reconstruction and re-simulation for training and evaluation, yet most methods for dynamic driving scenes still rely on per-scene optimization, known camera calibration, or short frame windows, making them slow and impractical. We revisit this problem from a feedforward perspective and introduce Driving Gaussian Grounded Transformer (DGGT), a unified framework for pose-free dynamic scene reconstruction. We note that the existing formulations, treating camera pose as a required input, limit flexibility and scalability. Instead, we reformulate pose as an output of the model, enabling reconstruction directly from sparse, unposed images and supporting an arbitrary number of views for long sequences. Our approach jointly predicts per-frame 3D Gaussian maps and camera parameters, disentangles dynamics with a lightweight dynamic head, and preserves temporal consistency with a lifespan head that modulates visibility over time. A diffusion-based rendering refinement further reduces motion/interpolation artifacts and improves novel-view quality under sparse inputs. The result is a single-pass, pose-free algorithm that achieves state-of-the-art performance and speed. Trained and evaluated on large-scale driving benchmarks (Waymo, nuScenes, Argoverse2), our method outperforms prior work both when trained on each dataset and in zero-shot transfer across datasets, and it scales well as the number of input frames increases.