We present a fast, robust and high-quality technique to skin a mesh with reference to a skeleton. We consider the space of possible skeleton deformations (based on skeletal constraints, or skeletal animations), and compute skinning weights based on an optimization scheme to obtain as-rigid-as-possible (ARAP) corresponding mesh deformations. We support stretchable-and-twistable bones (STBs) and spines by generalizing the ARAP deformations to stretchable deformers. In addition, our approach can optimize joint placements. If wanted, a user can guide and interact with the results, which is facilitated by an interactive feedback, reached via an efficient sparsification scheme. We demonstrate our technique on challenging inputs (STBs and spines, triangle and tetrahedral meshes featuring missing elements, boundaries, self-intersections or wire edges). @en

Mean Shift is a well-known clustering algorithm that has attractive properties such as the ability to find non convex and local clusters even in high dimensional spaces, while remaining relatively insensitive to outliers. However, due to its poor computational performance, real-world applications are limited. In this article, we propose a novel acceleration strategy for the traditional Mean Shift algorithm, along with a two-layer strategy, resulting in a considerable performance increase, while maintaining high cluster quality.We also show how to to find clusters in a streaming environment with bounded memory, in which queries need to be answered at interactive rates, and for which no mean shift-based algorithm currently exists. Our online structure is updated at very minimal cost and as infrequently as possible, and we show how to detect the time at which an update needs to be triggered. Our technique is validated extensively in both static and streaming environments.@en

Performance capture systems are used to acquire high-quality animated 3D surfaces, usually in form of a dense 3D triangle mesh. Extracting a more compact, yet faithful representation is often desirable, but existing solutions for animated sequences are surface-based, which leads to a limited approximation power in the case of extreme simplification. We introduce animated sphere-meshes, which are meshes indexing a set of animated spheres. Our solution is the first to output an animated volumetric structure to approximate animated 3D surfaces and optimizes for the sphere approximation, connectivity, and temporal coherence. As a result, our algorithm produces a multi-resolution structure from which a level of simplification can be selected in real-time, preserving a faithful approximation of the input, even at the coarsest levels. We demonstrate the use of animated sphere-meshes for low-cost approximate collision detection. Additionally, we propose a skinning decomposition, which automatically rigs the input mesh to the chosen level of detail. The resulting set of weights are smooth, compress the animation, and enable easy edits.@en

We propose a novel framework for photometric stereo (PS) under low-light conditions using uncalibrated near-light illumination. It operates on free-form video sequences captured with a minimalistic and affordable setup. We address issues such as albedo variations, shadowing, perspective projections, and camera noise. Our method uses specular spheres detected with a perspective-correcting Hough transform to robustly triangulate light positions in the presence of outliers via a least-squares approach. Furthermore, we propose an iterative reweighting scheme in combination with an ℓp-norm minimizer to robustly solve the calibrated near-light PS problem. In contrast to other approaches, our framework reconstructs depth, albedo (relative to light source intensity), and normals simultaneously and is demonstrated on synthetic and real-world scenes.@en

Voxel-based approaches are today's standard to encode volume data. Recently, directed acyclic graphs (DAGs) were successfully used for compressing sparse voxel scenes as well, but they are restricted to a single bit of (geometry) information per voxel. We present a method to compress arbitrary data, such as colors, normals, or reflectance information. By decoupling geometry and voxel data via a novel mapping scheme, we are able to apply the DAG principle to encode the topology, while using a palette-based compression for the voxel attributes, leading to a drastic memory reduction. Our method outperforms existing state-of-the-art techniques and is well-suited for GPU architectures. We achieve real-time performance on commodity hardware for colored scenes with up to 17 hierarchical levels (a 128K3voxel resolution), which are stored fully in core. @en