Can I read DINeMo: Learning Neural Mesh Models with no 3D Annotations on EtoBox?

DINeMo: Learning Neural Mesh Models with no 3D Annotations by Guo, Weijie; Zhang, Guofeng; Ma, Wufei; Yuille, Alan is a scholarly article available to read on EtoBox.

What is DINeMo: Learning Neural Mesh Models with no 3D Annotations about?

Category-level 3D/6D pose estimation is a crucial step towards comprehensive 3D scene understanding, which would enable a broad range of applications in robotics and embodied AI. Recent works explored neural mesh models that approach a range of 2D and 3D tasks from an analysis-by-synthesis perspective. Despite the largely enhanced robustness to partial occlusion and domain shifts, these methods depended heavily on 3D annotations for part-contrastive learning, which confines them to a narrow set of categories and hinders efficient scaling. In this work, we present DINeMo, a novel neural mesh model that is trained with no 3D annotations by leveraging pseudo-correspondence obtained from large visual foundation models. We adopt a bidirectional pseudo-correspondence generation method, which produce pseudo correspondence utilize both local appearance features and global context information. Experimental results on car datasets demonstrate that our DINeMo outperforms previous zero- and few-shot 3D pose estimation by a wide margin, narrowing the gap with fully-supervised methods by 67.3%. Our DINeMo also scales effectively and efficiently when incorporating more unlabeled images during tra

Author

Guo, Weijie; Zhang, Guofeng; Ma, Wufei; Yuille, Alan

Published

2025

Language

EN