TU Wien Research Data
Not a member yet
512 research outputs found
Sort by
Materials of Erasmus+ TrainRDM Open Science "Early stage Researchers" Training Week
No full text<p>Presentations and exercises of the <a href="https://rdmtraininghub.eu/">TrainRDM</a> Open Science "Early stage Researchers" Training Week. The event was held at La Sapienza University of Rome, from 12th to 16th September 2022 .</p>
RT-Percept Sun Temple
No full text<p>Pre-rendered dataset used in <a href="https://jaliborc.github.io/rt-percept/">Training and Predicting Visual Error for Real-Time Applications</a> for the Sun Temple scene. Generated using the RT-Percept renderer and the RT-Percept scenes.</p>
RT-Percept Emerald Square
No full text<p>Pre-rendered dataset used in <a href="https://jaliborc.github.io/rt-percept/">Training and Predicting Visual Error for Real-Time Applications</a> for the Emerald Square scenes. Generated using the RT-Percept renderer and the RT-Percept scenes.</p>
Work-Based Learning in Manufacturing Industry
No full text<p>This dataset has been created by the IALF Working Group of Work-Based Learning. It consists of a meta analysis of work-based learning approaches in manufacturing, classified by type of work-based learning and manufacturing sector after NACE 2.</p>
Modelled emissions, river loads and river concentrations for PFOA and PFOS in 2016/2017 in Austrian surface waters
No full text<p>The data set contains the emission per pathway into surface waters, the resulting annual river loads and the mean annual concentrations for the substances perfluorooctanoic acid (PFOA, CAS-No. 335-67-1) and perfluorooctanesulfonic acid (PFOS, CAS-No. 1763-23-1) for the years 2016 and 2017 modelled with the MoRE model (adapted from Fuchs et al. 2017) for Austria and some neighboring catchments.</p><p>Three variants of model results are available in the data set: The Median model variant giving the most probable values, based on median values for the model input data and a Best-case and Worst-case variant which are based on the first and third quartile of the input monitoring data.</p><p>The spatial resolution of the data are hydrologic catchments with an average size of 105 km². Emissions are sums of emissions for the respective catchment. River loads and river concentrations are valid for the main river at the respective catchment outlet (tributaries and main river within the catchments may have higher or lower concentrations).</p><p>CSV-files use semicolon as field delimiter and comma as decimal separator. The catchments are supplied as geodata in format of ESRI-Shapefile and Geography Markup Language (GML). The coordinate reference system is ETRS_1989_LAEA (EPSG-Code: 3035).</p>
GigaDepth: Learning Depth from Structured Light with Branching Neural Networks
No full text<p>Structured light-based depth sensors provide accurate depth information independently of the scene appearance by extracting pattern positions from the captured pixel intensities.</p><p>Spatial neighborhood encoding, in particular, is a popular structured light approach for off-the-shelf hardware. However, it suffers from the distortion and fragmentation of the projected pattern by the scene's geometry in the vicinity of a pixel. This forces algorithms to find a delicate balance between depth prediction accuracy and robustness to pattern fragmentation or appearance change. While stereo matching provides more robustness at the expense of accuracy, we show that learning to regress a pixel's position within the projected pattern is not only more accurate when combined with classification but can be made equally robust. We propose to split the regression problem into smaller classification sub-problems in a coarse-to-fine manner with the use of a weight-adaptive layer that efficiently implements branching per-pixel Multilayer Perceptrons applied to features extracted by a Convolutional Neural Network.</p><p>As our approach requires full supervision, we train our algorithm on a rendered dataset sufficiently close to the real-world domain. On a separately captured real-world dataset, we show that our network outperforms state-of-the-art and is significantly more robust than other regression-based approaches.</p><ul><li><strong>dataset.zip</strong>: Training and validation data, synthetic as well as captured.<br> </li><li><strong>dataset_jpg.zip</strong>: The same dataset but with lossy <i>jpg</i> encoding for IR images.<br> </li><li><strong>trained_models.zip</strong>: Models trained on the synthetic data.</li></ul>
MSDv1: Manufacturing Sound Dataset for Classification of Work-related Actions and their Sound
No full text<p>Currently, a dataset that contains sounds specific to manufacturing and/or assembly environment is not available. Within the Master's thesis "<a href="https://repositum.tuwien.at/handle/20.500.12708/137036"><strong>Sound is Context: Acoustic Work Step Classification using Deep Learning</strong></a>", a suitable dataset was created and open-sourced to public and research. The dataset contains a selection of typical production activities, which can be classified based only on their typical sounds. The recordings were performed in the Pilot Factory of the Vienna University of Technology.</p><p>The chosen production activities/sounds are:</p><ul><li>Grabing screws from a box</li><li>Sanding</li><li>Filing</li><li>Hammer</li><li>Cordless screwdriver</li><li>Press drill</li><li>Bench grinder</li></ul><p>The recording devices were the built-in microphone of a iPhone 13 Mini Smartphone and the Apple AUX-Headset microphone on a ASUS ZenBook 14 UM433I laptop. As smartphone recording applications, the standard Voice Memos App from Apple and the Voice Recorder & Memos Pro App from Linfei Ltd. were used, and on the laptop the standard Windows Voice Recorder was used. The different recording positions were 80cm away from the noise source, close (meaning 0-30cm), at the chest pocket, on a shelve above, reverse (meaning microphone was facing away or had an obstacle in between), or mixed (which means a variety of the before mentioned positions). To produce the sounds, the tempo, the intensity, the movement pattern, the rhythm, the hand, etc. were changed regularly, furthermore, due to the current work operation in the pilot factory, a wide variety of background noises were also recorded, which additionally increases the variance.</p><p>The recordings are available raw, without any pre-processing. Please refer to the thesis for information about pre-processing and the classification model.</p>
Untere Traisenbacher Höhle (UTB_104423)
No full text<p> Untere Traisenbacher Höhle - terrestrial scan no. UTB_104423. Cleaned and partly downsampled version (approx. 800K points). Acquired with Riegl VZ2000. Acquisition platform: tripod. </p>
FAIR for Sensitive Data
Get PDF<p>Materials (presentations and video) of the online webinar <strong>FAIR for Sensitive Data,</strong> organized by the FAIR Office Austria on March 23, 2022.<br><br>The goal of the webinar was to inform researchers on technical and legal aspects and hands-on practices when working with sensitive data. The workshop introduced tools that can be used for data anonymization, privacy-preserving processing, generation of synthetic data, etc. It included demos of selected tools for privacy-preserving data mining/analysis. Furthermore, the participants learnt about the <a href="https://www.sba-research.org/research/projects/wellfort/">WellFort</a> platform that integrates these tools and provides novel means for consent management.</p>
I Can Tell by Your Eyes! Continuous Gaze-Based Turn-Activity Prediction Reveals Spatial Familiarity
No full text<p>The data used for the analysis in the paper entitled "<strong>I Can Tell by Your Eyes! Continuous Gaze-Based Turn-Activity Prediction Reveals Spatial Familiarity</strong>" published in LIPIcs, Volume 240, COSIT 2022</p>
<h2><strong>How to Cite?</strong></h2>
<p><a href="https://drops.dagstuhl.de/entities/document/10.4230/LIPIcs.COSIT.2022.2">Alinaghi, N., Kattenbeck, M., & Giannopoulos, I. (2022). I can tell by your eyes! continuous gaze-based turn-activity prediction reveals spatial familiarity. In <em>15th International Conference on Spatial Information Theory (COSIT 2022)</em>. Schloss-Dagstuhl-Leibniz Zentrum für Informatik.</a></p><p>Spatial familiarity plays an essential role in the wayfinding decision-making process. Recent findings in wayfinding activity recognition domain suggest that wayfinders' turning behavior at junctions is strongly influenced by their spatial familiarity. By continuously monitoring wayfinders' turning behavior as reflected in their eye movements during the decision-making period (i.e., immediately after an instruction is received until reaching the corresponding junction for which the instruction was given), we provide evidence that familiar and unfamiliar wayfinders can be distinguished. By applying a pre-trained XGBoost turning activity classifier on gaze data collected in a real-world wayfinding task with 33 participants, our results suggest that familiar and unfamiliar wayfinders show different onset and intensity of turning behavior. These variations are not only present between the two classes -familiar vs. unfamiliar- but also within each class. The differences in turning-behavior within each class may stem from multiple sources, including different levels of familiarity with the environment.</p>