Abstract: Recently, self-supervised 6D object pose estimation, where synthetic images with object poses (sometimes jointly with un-annotated real images) are used for training, has attracted much ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results