Domain Transfer for 3D Pose Estimation from Color Images without Manual Annotations

Mahdi Rad (1) , Markus Oberweger (1) , Vincent Lepetit (2, 3)

1 ICG - Institute for Computer Graphics and Vision [Graz]
2 MANAO - Melting the frontiers between Light, Shape and Matter
3 LaBRI - Laboratoire Bordelais de Recherche en Informatique

We introduce a novel learning method for 3D pose estimation from color images. While acquiring annotations for color images is a difficult task, our approach circumvents this problem by learning a mapping from paired color and depth images captured with an RGB-D camera. We jointly learn the pose from synthetic depth images that are easy to generate, and learn to align these synthetic depth images with the real depth images. We show our approach for the task of 3D hand pose estimation and 3D object pose estimation, both from color images only. Our method achieves performances comparable to state-of-the-art methods on popular benchmark datasets, without requiring any annotations for the color images.

Domaines: Vision par ordinateur et reconnaissance de formes [cs.CV]; Apprentissage [cs.LG]

Mahdi Rad, Markus Oberweger, Vincent Lepetit. Domain Transfer for 3D Pose Estimation from Color Images without Manual Annotations. ACCV, 2018, Perth, Australia. ⟨hal-02509403⟩