Search
Results
-
Proactively Screening Machine Learning Pipelines with ARGUSEYES
(2023)SIGMOD '23: Companion of the 2023 International Conference on Management of DataSoftware systems that learn from data with machine learning (ML) are ubiquitous. ML pipelines in these applications often suffer from a variety of data-related issues, such as data leakage, label errors or fairness violations, which require reasoning about complex dependencies between their inputs and outputs. These issues are usually only detected in hindsight after deployment, after they caused harm in production. We demonstrate ArgusEyes, ...Conference Paper -
UVDoc: Neural Grid-based Document Unwarping
(2023)SA '23: SIGGRAPH Asia 2023 Conference PapersRestoring the original, flat appearance of a printed document from casual photographs of bent and wrinkled pages is a common everyday problem. In this paper we propose a novel method for grid-based single-image document unwarping. Our method performs geometric distortion correction via a fully convolutional deep neural network that learns to predict the 3D grid mesh of the document and the corresponding 2D unwarping grid in a dual-task ...Conference Paper -
Machine Learning Algorithm for the Scansion of Old Saxon Poetry
(2023)La memoria digitale: forme del testo e organizzazione della conoscenza. Atti del XII Convegno Annuale AIUCDSeveral scholars designed tools to perform the automatic scansion of poetry in many languages, but none of these tools deal with Old Saxon or Old English. This project aims to be a first attempt to create a tool for these languages. We implemented a Bidirectional Long Short-Term Memory (BiLSTM) model to perform the automatic scansion of Old Saxon and Old English poems. Since this model uses supervised learning, we manually annotated the ...Conference Paper -
XRgonomics: Facilitating the Creation of Ergonomic 3D Interfaces
(2021)Proceedings of the 2021 CHI Conference on Human Factors in Computing Systems (CHI '21)Arm discomfort is a common issue in Cross Reality applications involving prolonged mid-air interaction. Solving this problem is difficult because of the lack of tools and guidelines for 3D user interface design. Therefore, we propose a method to make existing ergonomic metrics available to creators during design by estimating the interaction cost at each reachable position in the user's environment. We present XRgonomics, a toolkit to ...Conference Paper -
Hintergründe der Tragwiderstandsbemessung von Brettsperrholz im Brandfall
(2021)Tagungsband Von der Forschung zur Praxis Erweiterung der Möglichkeiten im modernen Holzbau – Erdbebensicherheit und Brettsperrholz Mai 2021Conference Paper -
Schreiben im Kollektiv
(2021)Unterstellte LeseschaftenTextarbeit ist einsam und schafft Gemeinschaften, sie vollzieht sich im Privaten und zielt auf einen öffentlichen Resonanzraum. Taugt das, was ich schreibe? Wer wird mich lesen und wie werde ich gelesen werden? Um mit den Unsicherheiten umzugehen, die sich mit dem Übergang von Geschriebenem aus dem privaten Probierfeld in vorläufig Endgültiges, sprich: Publiziertes, verbinden, greifen Schreibende nicht erst seit dem Aufkommen des modernen ...Conference Paper -
Auswählen, bearbeiten, adressieren
(2021)Unterstellte LeseschaftenLiterarische Texte unterliegen impliziten und expliziten (Be-)Wertungen, die formen, was wir unter ‚Literatur‘ verstehen. Maßgebliche Instanzen dieser Aushandlungen waren lange Zeit vor allem die kritische und die philologische Kommunikation. Bevor deren kritisches Besteck jedoch zum Einsatz kommt, sind Texte aus einer Vielzahl von Texten ausgewählt und diese Texte zu Büchern und Werken gemacht worden. Innerhalb des modernen literarischen ...Conference Paper -
USCT data challenge 2019
(2021)Proceedings of the International Workshop on Medical Ultrasound TomographyRecent years have witnessed the active development of scanning systems and reconstructionalgorithms for ultrasound computed tomography (USCT) with applications to breast imagingfor early cancer detection. Despite these advances in hardware and software development,we encounter the need for reference data to develop, test and compare different imagingmethods. With the goals of encouraging scientific exchange and collaborations, providingbenchmarks ...Conference Paper -
Learning to Relate Depth and Semantics for Unsupervised Domain Adaptation
(2021)2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)We present an approach for encoding visual task relationships to improve model performance in an Unsupervised Domain Adaptation (UDA) setting. Semantic segmentation and monocular depth estimation are shown to be complementary tasks; in a multi-task learning setting, a proper encoding of their relationships can further improve performance on both tasks. Motivated by this observation, we propose a novel Cross-Task Relation Layer (CTRL), ...Conference Paper -
Urban Growth Within a Finite Space: Case-study Muscat
(2019)During the last four and half decades of physical urbanization, Muscat, the capital city of Oman, has received a massive influx of local and international migrants. Based on the National Centre of Statistics and Information (NCSI) in April 2016, Oman’s population was recorded at 4,441,448 with the population of Muscat Governorate forming more than half of Oman’s total population (2,395,412). According the NCSI statistics in 2010, Muscat ...Conference Paper