Christian Theobalt

TL;DR: This work presents an approach to add true fine-scale spatio-temporal shape detail to dynamic scene geometry captured from multi-view video footage and uses weak temporal priors on lighting, albedo and geometry which improve reconstruction quality yet allow for temporal variations in the data.

TL;DR: A new performance capture approach that incorporates a physically-based cloth model to reconstruct a rigged fully-animatable virtual double of a real person in loose apparel from multi-view video recordings and can now also create new real-time animations of actors captured in general apparel.

TL;DR: The key component of FaceVR is a robust algorithm to perform real-time facial motion capture of an actor who is wearing a head-mounted display (HMD), as well as a new data-driven approach for eye tracking from monocular videos.

TL;DR: A novel deep learning approach for monocular dense human performance capture trained in a weakly supervised manner based on multi-view supervision completely removing the need for training data with 3D ground truth annotations.

TL;DR: A fully automatic algorithm that jointly creates a rigged actor model commonly used for animation – skeleton, volumetric shape, appearance, and optionally a body surface and estimates the actor’s motion from multi-view video input only and optimizes the pose and shape parameters of this model fully automatically in a spatiotemporal way.