NVIDIA 开源 EgoHand：第一人称视频单目 3D 手部重建模型
原标题：EgoHand: Monocular 3D Hand Reconstruction from Egocentric Video
AI 摘要
NVIDIA 提出 EgoHand，一个面向第一人称视频的单目 3D 手部重建模型，使用 1400 万张图像（合成数据、野外视频、多相机动作捕捉）训练，采用 SOMA-Hand 参数化表示和 transformer 编码器，并引入相机线索与辅助 2D 分支。在 HOT3D、ARCTIC、H2O 上的零样本评测显示其关节位置、根平移和加速度误差低于对比基线，推理延迟约 9.01 ms，用户偏好研究也优于 Apple EgoDex 和 Ropedia Xperience-10M。模型与源码以 OpenMDW 许可开源，允许免费商用。
正文节选
largesymbols"00 largesymbols"01 EgoHand: Monocular 3D Hand Reconstruction from Egocentric Video Abstract We introduce EgoHand, a monocular 3D hand reconstruction model that achieves state-of-the-art performance on egocentric video. Central to our approach is the curation of a high-quality training corpus of 14 million images, combining synthetic data, in-the-wild video, and multi-camera motion capture to balance scene diversity and supervision quality. The model adopts SOMA-Hand as its parametri