HOReeNet: 3D-aware Hand-Object Grasping Reenactment
This addresses the problem of realistic hand-object interaction synthesis for applications like virtual reality or robotics, though it is incremental as it builds on existing reenactment and image translation methods.
The paper tackles the novel task of manipulating images to transfer objects and adjust 3D hand postures for tight grasping in 2D images, achieving state-of-the-art results on hand-object interaction datasets.
We present HOReeNet, which tackles the novel task of manipulating images involving hands, objects, and their interactions. Especially, we are interested in transferring objects of source images to target images and manipulating 3D hand postures to tightly grasp the transferred objects. Furthermore, the manipulation needs to be reflected in the 2D image space. In our reenactment scenario involving hand-object interactions, 3D reconstruction becomes essential as 3D contact reasoning between hands and objects is required to achieve a tight grasp. At the same time, to obtain high-quality 2D images from 3D space, well-designed 3D-to-2D projection and image refinement are required. Our HOReeNet is the first fully differentiable framework proposed for such a task. On hand-object interaction datasets, we compared our HOReeNet to the conventional image translation algorithms and reenactment algorithm. We demonstrated that our approach could achieved the state-of-the-art on the proposed task.