SLARM：面向动态场景的流式与语言对齐重建模型
SLARM: Streaming and Language-Aligned Reconstruction Model for Dynamic Scenes
SLARM 将语言对齐的语义特征与动态重建紧耦合，并采用窗口因果注意力实现低延迟流式推理，为动态场景重建提供了无需光流监督的高阶运动建模方案，可直接用于需要实时语义查询的 3D 重建任务。
原文依据 SLARM captures complex, non-uniform motion through higher-order motion modeling, trained solely on differentiable renderings without any flow supervision.
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论我们提出 SLARM，一个统一动态场景重建、语义理解和实时流式推理的前馈模型。SLARM 通过高阶运动建模捕捉复杂的非均匀运动，仅使用可微渲染进行训练，无需任何光流监督。此外，SLARM 从 LSeg 中蒸馏语义特征以获得语言对齐的表征。这一设计支持通过自然语言进行语义查询，并且语义与几何之间的紧密耦合进一步提升了动态重建的准确性和鲁棒性。同时，SLARM 使用基于窗口的因果注意力处理图像序列，实现稳定、低延迟的流式推理，且不会累积内存开销。在该统一框架内，SLARM 在动态估计、渲染质量和场景解析方面均达到最优结果，相比现有方法，运动精度提升 21%，重建峰值信噪比提升 1.6 dB，分割平均交并比提升 20%。
英文原文摘要 989 字符
We propose SLARM, a feed-forward model that unifies dynamic scene reconstruction, semantic understanding, and real-time streaming inference. SLARM captures complex, non-uniform motion through higher-order motion modeling, trained solely on differentiable renderings without any flow supervision. Besides, SLARM distills semantic features from LSeg to obtain language-aligned representations. This design enables semantic querying via natural language, and the tight coupling between semantics and geometry further enhances the accuracy and robustness of dynamic reconstruction. Moreover, SLARM processes image sequences using window-based causal attention, achieving stable, low-latency streaming inference without accumulating memory cost. Within this unified framework, SLARM achieves state-of-the-art results in dynamic estimation, rendering quality, and scene parsing, improving motion accuracy by 21%, reconstruction PSNR by 1.6 dB, and segmentation mIoU by 20% over existing methods.