We propose a fully-automatic method that computes from scratch point-to-point dense correspondences between isometric shapes under topological noise. While relying on pairwise distance preservation constraints is common and generally sufficient to handle isometric deformations, presence of topological noise needs further actions that we present as our main contributions. First, instead of comparing distances over two paths on two input surfaces, we cast fuzzy votes at the path endpoints based on topologically-robust heat diffusion from path vertices. Second, we make the matching even more stable to topological noise by introducing the so-called reodesics, which are locally shortest geodesics that go through robust matches. In addition to the five standard datasets for isometric shape correspondence with and without topological noise, we employ and release a sixth one geared specifically towards topological noise evaluation with ground-truth information. We demonstrate our qualitative and quantitative advantages over seven recent state-of-the-art methods on these six datasets.

As we move further into a spatial future, the demand for content creation tools is immense. Social media, gaming platforms and e-commerce have been converging into interactive spaces that involve spatial representations of the world the viewer is occupying, including digital humans. Creating digital representations of humans, or holograms, can be achieved through volumetric capture technologies. This method of bringing people to virtual three-dimensional environments has been rapidly increasing in popularity, but still lacks a key element: interactivity. In this paper we describe our work on producing interactive volumetric video that responds to viewers’ actions in real-time. We present the Fushimi Inari project: a commercial use case pushing the boundaries of what can be achieved with volumetric video and describe how our spatial content creation tools allow for interactive films to be created. Our contributions include blending volumetric clips, skeletonizing captures and applying multi-bone retargeting. We also provide means to integrate this in game engines for real-time photorealistic and interactive stories to be enjoyed by any viewer.