We propose an algorithm for real-time 6DOF pose tracking of rigid 3D objects using a monocular RGB camera. The key idea is to derive a region-based cost function using temporally consistent local color histograms. While such region-based cost functions are commonly optimized using first-order gradient descent techniques, we systematically derive a Gauss-Newton optimization scheme which gives rise to drastically faster convergence and highly accurate and robust tracking performance. We furthermore propose a novel complex dataset dedicated for the task of monocular object pose tracking and make it publicly available to the community. To our knowledge, it is the first to address the common and important scenario in which both the camera as well as the objects are moving simultaneously in cluttered scenes. In numerous experiments - including our own proposed dataset - we demonstrate that the proposed Gauss-Newton approach outperforms existing approaches, in particular in the presence of cluttered backgrounds, heterogeneous objects and partial occlusions.

We present a novel extension to embedding-based knowledge graph completion models which enables them to perform open-world link prediction, i.e. to predict facts for entities unseen in training based on their textual description. Our model combines a regular link prediction model learned from a knowledge graph with word embeddings learned from a textual corpus. After training both independently, we learn a transformation to map the embeddings of an entity’s name and description to the graph-based embedding space.In experiments on several datasets including FB20k, DBPedia50k and our new dataset FB15k-237-OWE, we demonstrate competitive results. Particularly, our approach exploits the full knowledge graph structure even when textual descriptions are scarce, does not require a joint training on graph and text, and can be applied to any embedding-based link prediction model, such as TransE, ComplEx and DistMult.

Facial animation on computationally weak systems is still mostly dependent on linear blendshape models. However, these models suffer from typical artifacts such as loss of volume, self-collisions, or erroneous soft tissue elasticity. In addition, while extensive effort is required to personalize blendshapes, there are limited options to simulate or manipulate physical and anatomical properties once a model has been crafted. Finally, second-order dynamics can only be represented to a limited extent. For decades, physics-based facial animation has been investigated as an alternative to linear blendshapes but is still cumbersome to deploy and results in high computational cost at runtime. We propose SoftDECA, an approach that provides the benefits of physics-based simulation while being as effortless and fast to use as linear blendshapes. SoftDECA is a novel hypernetwork that efficiently approximates a FEM-based facial simulation while generalizing over the comprehensive DECA model of human identities, facial expressions, and a wide range of material properties that can be locally adjusted without re-training. Along with SoftDECA, we introduce a pipeline for creating the needed high-resolution training data. Part of this pipeline is a novel layered head model that densely positions the biomechanical anatomy within a skin surface while avoiding self-intersections.

Purpose Computed tomography (CT) and in particular cone beam computed tomography (CBCT) have been increasingly used as a diagnostic tool in recent years. Patient motion during acquisition is common in CBCT due to long scan times. This results in degraded image quality and may potentially increase the number of retakes. Our aim was to develop a marker‐free iterative motion correction algorithm that works on the projection images and is suitable for local tomography. Methods We present an iterative motion correction algorithm that allows the patient's motion to be detected and taken into account during reconstruction. The core of our method is a fast GPU‐accelerated 3D reconstruction algorithm. Assuming rigid motion, motion correction is performed by minimizing a pixel‐wise cost function between all captured x‐ray images and parameterized projections of the reconstructed volume. Results Our method is marker‐free and requires only projection images. Furthermore, it can deal with local tomography data. We demonstrate the effectiveness of our approach on both simulated and real motion‐beset patient images. The results show that our new motion correction algorithm leads to accurate reconstructions with sharper edges, better contrasts and more detail. Conclusions The presented method allows for correction of patient motion with observable improvements in image quality compared to uncorrected reconstructions. Potentially, this may reduce the number of retakes caused by corrupted reconstructions due to patient movements.

Nowadays, cars offer many possibilities to explore the world around you by providing location-based information displayed on a 2D-Map. However, this information is often only available to front-seat passengers while being restricted to in-car displays. To propose a more natural way of interacting with the environment, we implemented an augmented reality head-mounted display to overlay points of interest onto the real world. We aim to compare multiple selection techniques for digital objects located outside a moving car by investigating head gaze with dwell time, head gaze with hardware button, eye gaze with hardware button, and hand pointing with gesture confirmation. Our study was conducted in a moving car under real-world conditions (N=22), with significant results indicating that hand pointing usage led to slower and less precise content selection while eye gaze was preferred by participants and performed on par with the other techniques.