Andreas Kriegler

Symmetric objects are common in daily life and industry, yet their inherent orientation ambiguities that impede the training of deep learning networks for pose estimation are rarely discussed in the literature. To cope with these ambiguities, existing solutions typically require the design of specific loss functions and network architectures or resort to symmetry-invariant evaluation metrics. In contrast, we focus on the numeric representation of the rotation itself, modifying trigonometric identities with the degrees of symmetry derived from the objects’ shapes. We use our representation, SARR, to obtain canonic (symmetry-resolved) poses for the symmetric objects in two popular 6D pose estimation datasets, T-LESS and ITODD, where SARR is unique and continuous w.r.t. the visual appearance. This allows us to use a standard CNN for 3D orientation estimation whose performance is evaluated with the symmetry-sensitive cosine distance AR_C. Our networks outperform the state of the art using AR_C and achieve satisfactory performance when using conventional symmetry-invariant measures. Our method does not require any 3D models but only depth, or, as part of an additional experiment, texture-less RGB/grayscale images as input. We also show that networks trained on SARR outperform the same networks trained on rotation matrices, Euler angles, quaternions, standard trigonometrics or the recently popular 6d representation – even in inference scenarios where no prior knowledge of the objects’ symmetry properties is available. Code and a visualization toolkit are available at https://github.com/akriegler/SARR.

This review paper provides an overview of papers that have studied body movement mirroring and synchrony within the field of human-robot interaction. The papers included in this review cover system studies, which focus on evaluating the technical aspects of mirroring and synchrony robotic systems, and user studies, which focus on measuring particular interaction outcomes or attitudes towards robots expressing mirroring and synchrony behaviors. We review the papers in terms of the employed robotic platforms and the focus on parts of the body, the techniques used to sense and react to human motion, the evaluation methods, the intended applications of the human-robot interaction systems and the scenarios utilized in user studies. Finally, challenges and possible future directions are considered and discussed.

A considerable amount of research is concerned with the challenging task of estimating three-dimensional (3D) pose and size for multi-object indoor scene con¯gurations. Many existing models rely on a priori known object models, such as 3D CAD models and are therefore limited to a prede¯ned set of object categories. This closed-set constraint limits the range of applications for robots interacting in dynamic environments where previously unseen objects may appear. This paper addresses this problem with a highly generic 3D bounding box detection method that relies entirely on geometric cues obtained from depth data percepts. While the generation of synthetic data, e.g. synthetic depth maps, is commonly used for this task, the well-known synth-to-real gap often emerges, which prohibits transition of models trained solely on synthetic data to the real world. To ameliorate this problem, we use stereo depth computation on synthetic data to obtain pseudo-realistic disparity maps. We then propose an intermediate representation, namely disparity-scaled surface normal (SN) images, which encodes geometry and at the same time preserves depth/scale information unlike the commonly used standard SNs. In a series of experiments, we demonstrate the usefulness of our approach, detecting everyday objects on a captured data set of tabletop scenes, and compare it to the popular PoseCNN model. We quantitatively show that standard SNs are less adequate for challenging 3D detection tasks by comparing predictions from the model trained on disparity alone, SNs and disparity-scaled SNs. Additionally, in an ablation study we investigate the minimal number of training samples required for such a learning task. Lastly, we make the tool used for 3D object annotation publicly available at: https://preview.tinyurl.com/3ycn8v5k. A video showcasing our results can be found at: https://preview.tinyurl.com/dzdzabek.

Enhancing logistics vehicles on airport aprons with assistant and autonomous capabilities offers the potential to significantly increase safety and efficiency of operations. However, this research area is still underrepresented compared to other automotive domains, especially regarding available image data, which is essential for training and benchmarking AI-based approaches. To mitigate this gap, we introduce a novel dataset specialized on static and dynamic objects commonly encountered while navigating apron areas. We propose an efficient approach for image acquisition as well as annotation of object instances and environmental parameters. Furthermore, we derive multiple dataset variants on which we conduct baseline classification and detection experiments. The resulting models are evaluated with respect to their overall performance and robustness against specific environmental conditions. The results are quite promising for future applications and provide essential insights regarding the selection of aggregation strategies as well as current potentials and limitations of similar approaches in this research domain.

Kuhn’s groundbreaking Structure divides scientific progress into four phases, the pre-paradigm period, normal science, scientific crisis and revolution. Most of the time a field advances incrementally, constrained and guided by a currently agreed upon paradigm. Creative phases emerge when phenomena occur which lack satisfactory explanation within the current paradigm (the crisis) until a new one replaces it (the revolution). This model of science was mainly laid out by exemplars from natural science, while we want to show that Kuhn’s work is also applicable for information sciences. We analyze the state of one field in particular, computer vision, using Kuhn’s vocabulary. Following significant technology-driven advances of machine learning methods in the age of deep learning, researchers in computer vision were eager to accept the models that now dominate the state of the art. We discuss the current state of the field especially in light of the deep learning revolution and conclude that current deep learning methods cannot fully constitute a paradigm for computer vision in the Kuhnian sense.

Despite their increasing demand for assistant and autonomous systems, the recent shift towards data-driven approaches has hardly reached aerial domains, partly due to a lack of specific training and test data. We introduce the Aircraft Context Dataset, a composition of two intercompatible large-scale and versatile image datasets focusing on manned aircraft and UAVs, respectively. In addition to fine-grained annotations for multiple learning tasks, we define and apply a set of relevant meta-parameters and showcase their potential to quantify dataset variability as well as the impact of environmental conditions on model performance. Baseline experiments are conducted for detection, classification and semantic labeling on multiple dataset variants. Their evaluation clearly shows that our contribution is an essential step towards overcoming the data gap and that the proposed variability concept significantly increases the efficiency of specializing models as well as continuously and purposefully extending the dataset.

For mobile robots to be considered autonomous they must reach target locations in required pose, a procedure referred to as docking. Popular current solutions use LiDARs combined with sizeable docking stations but these systems struggle by incorrectly detecting dynamic obstacles. This paper instead proposes a vision-based framework for docking a mobile robot. Faster R-CNN is used for detecting arbitrary visual markers. The pose of the robot is estimated using the solvePnP algorithm relating 2D-3D point pairs. Following exhaustive experiments, it is shown that solvePnP gives systematically inaccurate pose estimates in the x-axis pointing to the side. Pose estimates are off by ten to fifty centimeters and could therefore not be used for docking the robot. Insights are provided to circumvent similar problems in future applications.