[论文] F2G-Pose: Geometry-Aware Foundation Feature Lifting for Direct RGB-D C...

论文概要

研究领域: CV 作者: Jinyu Zhang, Haitao Lin, Jiashu Hou, Xiangyang Xue, Yanwei Fu 发布时间: 2026-10-22 arXiv: 2610.31623

中文摘要

类别级RGB-D物体位姿估计需在无实例CAD模型的情况下恢复3D旋转和平移，因类内变化和对称性而高度歧义。现有方法难以平衡可靠性（中间结构方法易误差传播）和效率（扩散模型推理延迟高）。F2G-Pose提出实时直接单遍框架，一次前馈即从单帧RGB-D预测位姿、度量尺寸和相机坐标系完整形状：将密集视觉基础特征提升到部分点云，DGCNN编码局部几何，混合专家Transformer整合物体级结构。仅在合成SOPE数据上训练即达到SOPE和ROPE最先进水平，28 FPS实时推理，并在HouseCat6D和HANDAL上展现强大泛化能力。

原文摘要

This paper studies category-level RGB-D object pose estimation, which recovers an object's 3D rotation and translation without instance-specific CAD models or reference views. Current solutions struggle to balance reliability and efficiency: methods based on intermediate structures are prone to error propagation, while diffusion models suffer from high inference latency. We present F2G-Pose, a real-time direct, single-pass framework that predicts category-level pose, metric size, and camera-frame completed shape from a single RGB-D observation in one feed-forward pass. F2G-Pose lifts dense visual foundation features to partial point clouds, encodes local geometry with DGCNN tokenization, and integrates object-level structure with a Mixture-of-Experts Transformer. Trained only on synthetic ...

#论文 #arXiv #CV #小凯