1,721,159 research outputs found
Learning Fluid Flow Visualizations From In-Flight Images With Tufts
No full text<p><strong>Abstract:</strong> To better understand fluid flows around aerial systems, strips of wire or rope, widely known as tufts, are often used to visualize the local flow direction. This letter presents a computer vision system that automatically extracts the shape of tufts from images, which have been collected during real flights of a helicopter and an unmanned aerial vehicle (UAV). As images from these aerial systems present challenges to both the model-based computer vision and the end-to-end supervised deep learning techniques, we propose a semantic segmentation pipeline that consists of three uncertainty-based modules namely, (a) active learning for object detection, (b) label propagation for object classification, and (c) weakly supervised instance segmentation. Overall, these probabilistic approaches facilitate the learning process without requiring any manual annotations of semantic segmentation masks. Empirically, we motivate our design choices through comparative assessments and provide real-world demonstrations of the proposed concept, for the first time to our knowledge.</p>
<p><strong>MetaInfo: </strong>This dataset accompanies the publication: Learning Fluid Flow Visualizations From In-Flight Images With Tufts at IEEE RA-L 2023. The data consists of images and annotations of tufts for fluid flow visualization, which were used to validate a semantic segmentation method based on uncertainty in deep learning.</p>
<p>Images of tufts from the DLR helicopter EC135-ACT⁄FHS have been collected at the Institute of Aerodynamics and Flow Technology. Images of tufts from the stratospheric flight of the robot HABLEG have been collected at the Institute of Robotics and Mechatronics.</p>
<p><strong>Project website:</strong> https://sites.google.com/view/tuftrecognition/</p>
<p>This website contains more details about the data collection procedures, and how tufts have been placed on the considered aerial vehicles.</p>
<p><strong>Bibtex:</strong> This dataset can be cited as the same way of the original article.</p>
<blockquote>
<pre>@ARTICLE{10109020,
author={Lee, Jongseok and Olsman, W.F.J. and Triebel, Rudolph},
journal={IEEE Robotics and Automation Letters},
title={Learning Fluid Flow Visualizations From In-Flight Images With Tufts},
year={2023},
volume={8},
number={6},
pages={3677-3684},
doi={10.1109/LRA.2023.3270746}}
</pre>
</blockquote>
Optimal intrinsic descriptors for non-rigid shape analysis
Full text linkWe propose novel point descriptors for 3D shapes with the potential to match two shapes representing the same object undergoing natural deformations. These deformations are more general than the often assumed isometries, and we use labeled training data to learn optimal descriptors for such cases. Furthermore, instead of explicitly defining the descriptor, we introduce new Mercer kernels, for which we formally show that their corresponding feature space mapping is a generalization of either the Heat Kernel Signature or the Wave Kernel Signature. I.e. the proposed descriptors are guaranteed to be at least as precise as any Heat Kernel Signature or Wave Kernel Signature of any parameterisation. In experiments, we show that our implicitly defined, infinite-dimensional descriptors can better deal with non-isometric deformations than state-of-the-art methods
GPGM-SLAM: a Robust SLAM System for Unstructured Planetary Environments with Gaussian Process Gradient Maps
Full text linkSimultaneous Localization and Mapping (SLAM) techniques play a key role towards long-term autonomy of mobile robots due to the ability to correct localization errors and produce consistent maps of an environment over time. Contrarily to urban or man-made environments, where the presence of unique objects and structures offer unique cues for localization, the apperance of
unstructured natural environments is often ambiguous and self-similar, hindering the performances of loop closure detection. In this paper, we present an approach to improve the robustness of place
recognition in the context of a submap-based stereo SLAM based on Gaussian Process Gradient Maps (GPGMaps). GPGMaps embed a continuous representation of the gradients of the local terrain
elevation by means of Gaussian Process regression and Structured Kernel Interpolation, given solely noisy elevation measurements. We leverage the imagelike structure of GPGMaps to detect loop
closures using traditional visual features and Bag of Words. GPGMap matching is performed as an SE(2) alignment to establish loop closure constraints within a pose graph. We evaluate the
proposed pipeline on a variety of datasets recorded on Mt. Etna, Sicily and in the Morocco desert, respectively Moon- and Mars-like environments, and we compare the localization performances with
state-of-the-art approaches for visual SLAM and visual loop closure detection
Deep visual human sensing with application in robotics
No full textThanks to advances in robotics, navigation, localization, and perception in the last decade, mobile robots (which includes self-driving cars) are recently starting to be deployed in everyday life scenarios, surrounded by people. In such situations, understanding the people surrounding the robot is crucial. This thesis consists of a collection of works which significantly advance the state of the art in visual understanding of humans. First, we introduce a fundamentally new paradigm for performing detection in 2D LiDAR scans. Our detector, dubbed DROW, is based on a voting scheme, where each individual measuring point casts a vote. It is completely data-driven, naturally multiclass, and outperforms previous detectors and even trackers significantly. Orientation of people, as well as their head orientation, are important higher-level cues for attention and motion prediction. We introduce a new neural network output module, the Biternions, and a corresponding von-Mises loss function, which allow for accurate, continuous orientation prediction using only weak, discrete labeling of data. We furthermore extend it with a principled, learned measure of confidence in its own prediction. Then, we take a closer look at learning semantic embeddings of images, with focus mainly on person re-identification, and promising results on object recognition. We demonstrate that triplet-loss based approaches perform much better than previously assumed, while being a simple and ideologically “clean” family of methods. In fact, our proposed model using ImageNet pre-trained ResNet50, batch-hard triplet loss, PK-batches, and a soft margin significantly outperforms the state-of-the-art on multiple person re-identification benchmarks, as well as on fine-grained car, bird, and product recognition benchmarks. All aforementioned advances make use of deep learning, which typically results in algorithms which require hardware accelerators on the robot. Having multiple such components on a single robot comes at a cost. We investigate ways of mitigating this cost in our DetTA pipeline which leverages a tracker to perform strided execution of analysis modules (thus significantly reducing load) and per-person smoothing of the results (thus not decreasing prediction accuracy). Finally, motivated by the importance of tracking on mobile robots and our strong person re-identification results, we investigate a completely novel formulation to tracking which makes use of a solid person re-identification model from the ground up, bypassing the need for complicated data-association. This new formulation goes one step further towards end-to-end learning of tracking and opens up many novel research opportunities
Kollaborative Lokalisierung und Kartenerstellung für die autonome planetare Exploration : Verteilter 6D SLAM für stereokamerabasierte Roboterteams in GNSS-freien Umgebungen
No full textMobile robots are a crucial element of present and future scientific missions to explore the surfaces of foreign celestial bodies such as Moon and Mars. The deployment of teams of robots allows to improve efficiency and robustness in such challenging environments. As long communication round-trip times to Earth render the teleoperation of robotic systems inefficient to impossible, on-board autonomy is a key to success. The robots operate in Global Navigation Satellite System (GNSS)-denied environments and thus have to rely on space-suitable on-board sensors such as stereo camera systems. They need to be able to localize themselves online, to model their surroundings, as well as to share information about the environment and their position therein. These capabilities constitute the basis for the local autonomy of each system as well as for any coordinated joint action within the team, such as collaborative autonomous exploration. In this thesis, we present a novel approach for stereo vision-based on-board and online Simultaneous Localization and Mapping (SLAM) for multi-robot teams given the challenges imposed by planetary exploration missions. We combine distributed local and decentralized global estimation methods to get the best of both worlds: A local reference filter on each robot provides real-time local state estimates required for robot control and fast reactive behaviors. We designed a novel graph topology to incorporate these state estimates into an online incremental graph optimization to compute global pose and map estimates that serve as input to higher-level autonomy functions. In order to model the 3D geometry of the environment, we generate dense 3D point cloud and probabilistic voxel-grid maps from noisy stereo data. We distribute the computational load and reduce the required communication bandwidth between robots by locally aggregating high-bandwidth vision data into partial maps that are then exchanged between robots and composed into global models of the environment. We developed methods for intra- and inter-robot map matching to recognize previously visited locations in semi- and unstructured environments based on their estimated local geometry, which is mostly invariant to light conditions as well as different sensors and viewpoints in heterogeneous multi-robot teams. A decoupling of observable and unobservable states in the local filter allows us to introduce a novel optimization: Enforcing all submaps to be gravity-aligned, we can reduce the dimensionality of the map matching from 6D to 4D. In addition to map matches, the robots use visual fiducial markers to detect each other. In this context, we present a novel method for modeling the errors of the loop closure transformations that are estimated from these detections. We demonstrate the robustness of our methods by integrating them on a total of five different ground-based and aerial mobile robots that were deployed in a total of 31 real-world experiments for quantitative evaluations in semi- and unstructured indoor and outdoor settings. In addition, we validated our SLAM framework through several different demonstrations at four public events in Moon and Mars-like environments. These include, among others, autonomous multi-robot exploration tests at a Moon-analogue site on top of the volcano Mt. Etna, Italy, as well as the collaborative mapping of a Mars-like environment with a heterogeneous robotic team of flying and driving robots in more than 35 public demonstration runs
Going Beyond Counting First Authors in Author Co-citation Analysis
Full text linkThe present study examines one of the fundamental aspects of author co-citation analysis (ACA) - the way co-citation
counts are defined. Co-citation counting provides the data on which all subsequent statistical analyses and mappings
are based, and we compare ACA results based on two different types of co-citation counting - the traditional type that
only counts the first one among a cited work's authors on the one hand and a non-traditional type that takes into
account the first 5 authors of a cited work on the other hand. Results indicate that the picture produced through this non-traditional author co-citation counting contains more coherent author groups and is therefore considerably clearer. However, this picture represents fewer specialties in the research field being studied than that produced through the traditional first-author co-citation counting when the same number of top-ranked authors is selected and analyzed. Reasons for these effects are discussed
Multi-Sensor and Multi-Modal Localization in Indoor Environments on Robotic Platforms
Full text linkThis dissertation presents a detailed exploration of multi-sensor visual odometry systems, with a focus on enhancing robustness in indoor environments. The core contribution lies in the development of an advanced ego-state estimation framework, wherein visual odometry is bolstered by multiple visual sensors to overcome common challenges such as occlusions and textureless surfaces. By addressing frequent loss-of-tracking (LoT) events, the system ensures continuous, reliable localization in complex, cluttered indoor settings, such as households and elderly care facilities.
To further augment situational awareness, sound source localization (SSL) is integrated as a complementary modality. Its fusion with visual data signicantly enhances the robot’s perception of the environment, enabling the detection and identication of objects and events that may be visually occluded or otherwise undetectable. This multi-modal fusion provides a more holistic understanding of the robot’s surroundings, contributing to improved operational reliability in dynamic, human-centered environments.
A key feature of this research is the introduction of IndoorMCD, a novel multisensor benchmark specically designed to evaluate localization performance in indoor environments. Additionally, this work introduces URSim, an online real-time visual simulation framework that enables rigorous testing of multisensor localization systems under various conditions. Extensive experimental
validation, using both real-world scenarios and simulated environments, demonstrates the robustness and fault tolerance of the proposed system.
This research advances the state-of-the-art in robotic perception and indoor localization by providing a multi-modal, fault-tolerant approach to localization, oering valuable contributions to both theoretical understanding and practical application in robotics
Variations on the Author
Full text link“Variations on the Author” discusses two of Eduardo Coutinho’s recent films (Um Dia na Vida, from 2010, and Últimas Conversas, posthumously released in 2015) and their contribution to the general question of documentary authorship. The director’s filmography is characterized by a consistent yet self-effacing form of authorial self-inscription: Coutinho often features as an interviewer that rather than express opinions propels discourses; an interviewer that is good at listening. This mode of self-inscription characterizes him as an author who is not expressive but who is nonetheless markedly present on the screen. In Um Dia na Vida, however, Coutinho is completely absent form the image, while Últimas Conversas, on the contrary, includes a confessional prologue that moves the director from the margins to the center of his films. This article examines the ways in which these works stand out in the filmography of a director who offers new insights into the notion of cinematic authorship
Appropriate Similarity Measures for Author Cocitation Analysis
Full text linkWe provide a number of new insights into the methodological discussion about author cocitation analysis. We first argue that the use of the Pearson correlation for measuring the similarity between authors’ cocitation profiles is not very satisfactory. We then discuss what kind of similarity measures may be used as an alternative to the Pearson correlation. We consider three similarity measures in particular. One is the well-known cosine. The other two similarity measures have not been used before in the bibliometric literature. Finally, we show by means of an example that our findings have a high practical relevance.information science;Pearson correlation;cosine;similarity measure;author cocitation analysis
- …
