MuM：面向三维视觉的多视图掩码图像建模
MuM: Multi-View Masked Image Modeling for 3D Vision
把 MAE 从双视图 CroCo 扩展到任意多视图、统一掩码加帧间注意力轻量解码器，为需要几何推理的视觉任务（前馈重建、稠密匹配、相对位姿）提供了一个可直接替换 DINOv3/CroCo 的预训练编码器方案。
原文依据 we extend MAE to arbitrarily many views of the same scene
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论图像上的自监督学习旨在从无标注数据中提取有意义的视觉表征。当扩展到大规模数据集时，这一范式已取得最优性能，由此训练出的 DINOv3 等模型得到了广泛采用。然而，以往大多数工作针对的是语义理解而非几何推理。一个重要的例外是跨视图补全（Cross-View Completion，CroCo），它是为三维理解量身定制的一种掩码自编码（MAE）形式。在本工作中，我们延续 CroCo 提出的路线，专注于学习面向三维视觉的特征。简而言之，我们将 MAE 扩展到同一场景的任意多个视图。通过对所有视图施加均匀掩码，并采用带帧间注意力的轻量解码器，我们的方法在本质上比 CroCo 更简单、更具可扩展性。我们在前馈重建、稠密图像匹配和相对位姿估计等下游任务上对所得模型 MuM 进行了广泛评测，发现它优于最优的视觉编码器 DINOv3 和 CroCo v2。
英文原文摘要 1097 字符
Self-supervised learning on images seeks to extract meaningful visual representations from unlabeled data. When scaled to large datasets, this paradigm has achieved state-of-the-art performance and the resulting trained models such as DINOv3 have seen widespread adoption. However, most prior efforts are optimized for semantic understanding rather than geometric reasoning. One important exception is Cross-View Completion, CroCo, which is a form of masked autoencoding (MAE) tailored for 3D understanding. In this work, we continue on the path proposed by CroCo and focus on learning features tailored for 3D vision. In a nutshell, we extend MAE to arbitrarily many views of the same scene. By uniformly masking all views and employing a lightweight decoder with inter-frame attention, our approach is inherently simpler and more scalable than CroCo. We evaluate the resulting model, MuM, extensively on downstream tasks including feedforward reconstruction, dense image matching and relative pose estimation, finding that it outperforms the state-of-the-art visual encoders DINOv3 and CroCo v2.