Contact-centric deformation learning

Fig. 1. We present a learning-based method to augment a subspace deformable simulation with contact-driven deformation detail. We learn contact deformations in a contact-centric manner, which allows us to significantly reduce the sampling of configurations of the deformable object, and subsequently learn highly complex deformations. For this real-time simulation of the MANO model [Romero et al. 2017] with dynamics, we used just one pose of the hand for training. Notice the accurate high-resolution deformations due to contact with a rigid object, highlighted in the zoom-ins.

We propose a novel method to machine-learn highly detailed, nonlinear contact deformations for real-time dynamic simulation. We depart from previous deformation-learning strategies, and model contact deformations in a contact-centric manner. This strategy shows excellent generalization with respect to the object's configuration space, and it allows for simple and accurate learning. We complement the contact-centric learning strategy with two additional key ingredients: learning a continuous vector field of contact deformations, instead of a discrete approximation; and sparsifying the mapping between the contact configuration and contact deformations. These two ingredients further contribute to the accuracy, efficiency, and generalization of the method. We integrate our learning-based contact deformation model with subspace dynamics, showing real-time dynamic simulations with fine contact deformation detail.