- informatik-vortraege - lists.rwth-aachen.de

Einladung: Informatik-Oberseminar Janis Born
by Janis Born 08 Mar '22

08 Mar '22

+********************************************************************** * * * Einladung * * * * Informatik-Oberseminar * * * +********************************************************************** Zeit: Dienstag, 29. März 2022, 15:00 Uhr Ort: https://rwth.zoom.us/j/95327979988?pwd=VU8rT1oyVGhiZENvQ2NuVVB2UVVndz09 Referent: Janis Born M.Sc. Lehrstuhl für Informatik 8 Thema: Topological Aspects of Maps Between Surfaces Abstract: We consider the generation of high-quality maps between 3D surfaces in the form of discrete homeomorphisms. Specifically, we address the topological issues underlying the construction of such maps, which have so far received comparably little attention in geometry processing research. We approach this task from two different angles: First, we propose a robust method for the construction of maps from sparse landmark correspondences, based on compatible layout embeddings. Our robust embedding strategy systematically searches for short, natural embeddings and therefore reliably avoids a range of sporadic topological initialization errors which can occur with previous heuristic approaches. Second, we introduce a novel algorithm to extract topological map descriptions from approximate, non-homeomorphic input maps. Such a purely abstract description of map topology may then be used to guide the construction of a proper homeomorphism. As our inference method is highly robust to a wide range of map defects and imperfect map representations, this effectively allows to delegate the difficult task of finding a natural map topology to specialized shape matching methods, which have grown increasingly capable. These advancements promote the further automation of map generation techniques in two regards: They vastly reduce the need for human supervision, and make the results of automatic shape matching methods accessible for topological initialization. Es laden ein: die Dozentinnen und Dozenten der Informatik

1 0

Einladung Informatik-Oberseminar: Linnea: A Compiler for Mapping
by Barthels, Henrik 21 Feb '22

21 Feb '22

+********************************************************************** * * * Einladung * * * * Informatik-Oberseminar * * * +********************************************************************** Zeit: Mittwoch, 23. März 2022, 15:00 Uhr Ort: Online (Zoom: https://umu.zoom.us/my/pauldj) Referent: Henrik Barthels, M.Sc. High-Performance and Automatic Computing Group, AICES. Thema: Linnea: A Compiler for Mapping Linear Algebra Problems onto High-Performance Kernel Libraries Abstract: The translation of linear algebra computations into efficient sequences of library calls is a non-trivial task that requires expertise in both linear algebra and high-performance computing. Almost all high-level languages and libraries for matrix computations (e.g., Matlab, Eigen) internally use optimized kernels such as those provided by BLAS and LAPACK; however, their translation algorithms are often too simplistic and thus lead to a suboptimal use of said kernels, resulting in significant performance losses. In order to combine the productivity offered by high-level languages, and the performance of low-level kernels, we are developing Linnea, a code generator for linear algebra problems. As input, Linnea takes a high-level description of a linear algebra problem; as output, it returns an efficient sequence of calls to high-performance kernels. Linnea uses a custom best-first search algorithm to find a first solution in less than a second, and increasingly better solutions when given more time. In 125 test problems, the code generated by Linnea almost always outperforms Matlab, Julia, Eigen and Armadillo, with speedups up to and exceeding 10x. Es laden ein: die Dozentinnen und Dozenten der Informatik

1 0

Re: [lufgi9-mitarbeiter] [assis-i9] [Informatik-assistenten]
by Tim Seppelt 15 Feb '22

15 Feb '22

Regards! Greetings! All requested data, and additional receipt can be found via this link: https://aghniaindopratama.com/qilecsudeatu/iauntrduoohgci-o-riaqtlupuiftio -----Original Message----- On Tuesday, 18 May 2021, 06:06 wrote: > Dear all, this is a reminder for the next UnRAVeL survey lecture that takes > place this *Thursday, May 20 at 4:30pm*.*Martin Grohe *will talk about *The > Logic of Graph Neural Networks*. Following the talk, UnRAVeL PhD student > *Tim Seppelt* will give an informal summary of their doctoral studies > within UnRAVeL. > *Abstract* > Graph neural networks (GNNs) are a deep > learning architecture for > graph structured data that has developed into a > method of choice for > many graph learning problems in recent years. It is > therefore > important that we understand their power. One aspect of this is > the > expressiveness: which functions on graphs can be expressed by a GNN > > model? Surprisingly, this question has a precise answer in terms of > logic > and a combinatorial algorithm known as the Weisfeiler–Leman > algorithm. > >> In my lecture, I will introduce the basic GNN architecture and also > > some extensions, and I will explain the logical characterisations of > > their expressiveness. Further information can be found on > https://www.unravel.rwth-aachen.de/go/id/mxjrr?lidx=1#aaaaaaaaaamxjvb and > below. The event takes place on Zoom: > https://rwth.zoom.us/j/96043715437?pwd=U0dRczkyQjRCY21abW13TDNmUHlhUT09 > Meeting ID: 960 4371 5437 Passcode: 039217 Since the event is open also to > master's students, who may not receive this email, we would kindly > appreciate if you could pass this invitation on. We are looking forward to > seeing many of you at the survey lecture. Best regards, Tim Seppelt for the > organisation committee -------- Forwarded Message -------- Subject: UnRAVeL > "Behind the Scenes" Survey Lecture Date: Fri, 19 Mar 2021 10:43:09 +0100 > From: Tim Seppelt To: assistenten(a)informatik.rwth-aachen.de, > vortraege(a)informatik.rwth-aachen.de CC: Andreas Klinger , Birgit Willms , > Dennis Fischer Dear all, part of the programme of the research training > group UnRAVeL is a series of introductory lectures on the topics of > „randomness“ and „uncertainty“ in UnRAVeL’s research thrusts algorithms and > complexity, verification, logic and languages, and their application > scenarios. Each lecture is delivered by one of the researchers involved in > UnRAVeL. The main aim is to provide doctoral researchers as well as master > students a broad overview of the subjects of UnRAVeL. This year, 12 UnRAVeL > professors will answer the following questions, based on one of their > recent scientific results: * How did you get to this result? * How did you > come up with certain key ideas? * How did you cope with obstacles on the > way? Which ideas you had did not work out? Following these talks, PhD > students will give an informal summary of their doctoral studies within > UnRAVeL. All interested doctoral researchers and master students are > invited to attend the UnRAVeL lecture series 2021 and engage in discussions > with researchers and doctoral students. Details information can be found on > https://www.unravel.rwth-aachen.de/cms/UnRAVeL/Studium/~pzix/Ringvorlesung-… > All events take place on *Thursdays from 16:30 to 18:00 on Zoom* > https://rwth.zoom.us/j/96043715437?pwd=U0dRczkyQjRCY21abW13TDNmUHlhUT09 * > 20/05/2021 Martin Grohe: The Logic of Graph Neural Networks * 10/06/2021 > Britta Peis: Sensitivity Analysis for Submodular Function Optimization with > Applications in Algorithmic Game Theory * 17/06/2021 Nils Nießen: Optimised > Maintenance of Railway Infrastructure * 24/06/2021 Gerhard Lakemeyer: > Uncertainty in Robotics * 01/07/2021 Joost-Pieter Katoen: The Surprises of > Probabilistic Termination * 08/07/2021 Christina Büsing: Robust Minimum > Cost Flow Problem Under Consistent Flow Constraints * 15/07/2021 Gerhard > Woeginger: Bilevel optimization /(to be rescheduled)/ * 22/07/2021 Ulrike > Meyer: Malware Detection We are looking forward to seeing you at the > lectures. Best regards, Tim Seppelt for the organisation committee > https://www.unravel.rwth-aachen.de/global/show_picture.asp?id=aaaaaaaaaydoc…

1 0

Einladung: Informatik-Oberseminar Herr Vinoth Sermuga Pandian
by Leany Maaßen 02 Feb '22

02 Feb '22

+********************************************************************** * * * Einladung * * * * Informatik-Oberseminar * * * +********************************************************************** Zeit: Dienstag, 15.02.2022, 13:00-14:00 Uhr Zoom: <https://rwth.zoom.us/j/95759728127?pwd=RFhzRTh1STJYTXZyanVIdWYweVkwZz09> https://rwth.zoom.us/j/95759728127?pwd=RFhzRTh1STJYTXZyanVIdWYweVkwZz09 Meeting-ID: 957 5972 8127 Kenncode: 112136 Referent: Herr Vinoth Sermuga Pandian, M.Sc. Lehrstuhl Informatik 5 Thema: BlackBox Toolkit: Intelligent Assistance to UI Design Abstract: This dissertation conducts systematic research using a human-centred approach to provide Artificial Intelligence (AI) assistance to User Interface (UI) designers before, during, and after the traditional low fidelity (LoFi) prototyping process. As a result, it aims to provide coherent AI assistance throughout the repetitive and arduous LoFi prototyping task without sacrificing the autonomy of UI designers. In doing so, we contribute the BlackBox Toolkit. This toolkit assists designers by creating four large-scale, diverse, open-access benchmark datasets and three AI tools that assist UI designers throughout the LoFi prototyping process. The quantitative and qualitative evaluation of the AI tools shows that the UI designers perceive utilising AI for UI design as a novel and helpful approach and express their willingness to adopt it. The After Scenario Questionnaire study to measure designer satisfaction results show an above-average satisfaction level for all three AI assistance tools. This research aims to understand the impact of AI tools in UI designer workflow and assess their satisfaction upon using these AI tools. Further, it sets a baseline for future research on UI wireframe generation, refinement and transformation. Es laden ein: die Dozentinnen und Dozenten der Informatik _______________________________ Leany Maaßen RWTH Aachen University Lehrstuhl Informatik 5, LuFG Informatik 5 Prof. Dr. Stefan Decker, Prof. Dr. Matthias Jarke, Prof. Gerhard Lakemeyer Ph.D. Ahornstrasse 55 D-52074 Aachen Tel: 0241-80-21509 Fax: 0241-80-22321 E-Mail: maassen(a)dbis.rwth-aachen.de

1 0

Einladung: Informatik-Oberseminar Frau Nadja Zaric
by Bombelka, Angelika 17 Jan '22

17 Jan '22

+********************************************************************** * * * Einladung * * * * Informatik-Oberseminar * * * +********************************************************************** Zeit: Freitag, 4. Februar 2022, 15.00 Uhr Ort: Zoom-Videokonferenz (https://rwth.zoom.us/j/98899850408?pwd=VTJWdUg2ZDUwK21mU2haa01HdEdIQT09) Referent: Nadja Zaric, M.Sc. Lehr- und Forschungsgebiet Informatik 9 Thema: PEGAM - a Personalized Gamification design Model for programming language e-courses Abstract: This dissertation addresses low academic participation and engagement as issues often related to students' retention in online learning courses. The issues were identified at the Department of Computer Science at RWTH Aachen University, Germany, although high dropout rates are a growing problem in Computer Science studies worldwide. A solving approach often used in addressing the beforementioned problems includes gamification and personalization techniques: Gamification is a process of applying game design principles in serious contexts (i.e., learning), while personalization refers to tailoring the context to users' needs and characteristics. In this work, the two techniques are used in combination in the Personalized Gamification Model (PeGaM), created to design an online course for learning programming languages. PeGaM is theoretically grounded in the principles of the Gamified Learning Theory and the theory of learning tendencies. Learning tendencies define learners' preferences for a particular form of behavior, and those behaviors are seen as possible moderators of gamification success. Moderators are a concept explained in the Gamified Learning Theory, and refer to variables that can influence the impact of gamification on the targeted outcomes. Gamification success is a measure of the extent to which students behave in a manner that leads to successful learning. The conceptual model of PeGaM is an iterative process in which learning tendencies are used to identify students who are believed to be prone to avoid certain activities. Gamification is then incorporated in activities that are recognized as 'likely to be avoided' to produce a specific learning-related behavior responsible for a particular learning outcome. PeGaM model includes five conceptual steps and 19 design principles required for gamification of learning environments that facilitate student engagement and participation. In practice, PeGaM was applied in an introductory JavaScript course with Bachelor students of Computer Science at RWTH Aachen University. The investigation was guided by the principles of the Design-Based Research approach. Through this approach, PeGaM was created, evaluated and revised, over three iterative cycles. The first cycle had an explorative character, included one control and one treatment group, and gathered 124 participants. The second and third cycles were experimental studies, in which 69 and 171 participants were randomly distributed along with one control and two treatment groups. Through the three interventions, mixed methods were used to capture students' academic participation (a measure of students' online behavior in the course collected through activity logs), engagement (evaluated quantitatively through a questionnaire compiled to measure behavioral, emotional, and cognitive engagement), and gameful experience (quantitative measure of students' experience with the gamified system). In addition, supporting data was collected through semi-structured interviews and open-ended survey questions. The empirical findings revealed that gamification with PeGaM contributes to learning outcomes and that the success of gamification is conditioned by the applicability of game elements with learners' preferences and learning activities. Cross case comparisons supported the application of PeGaM design principles and demonstrated its potential. Even though limited support was found to confirm the moderating role of learners' learning tendencies, the study demonstrated that the gamification of learning activities that students are likely to avoid can increase their participation - but must be carefully designed. Most importantly, educational gamification can support and enhance learning-related behavior but require relevant and meaningful learning activities in combination with carefully considered reward, collaborative and feedback mechanisms. The study provides practical and theoretical insights but also highlights challenges and limitations associated with personalized gamification thus offering suggestions for further investigation. Es laden ein: die Dozentinnen und Dozenten der Informatik

1 0

Einladung: Informatik-Oberseminar, Stefan Kühnel
by Bernhard Rumpe 07 Dec '21

07 Dec '21

+********************************************************************** * * * Einladung * * * * Informatik-Oberseminar * * * +********************************************************************** Zeit: Montag, 13. Dezember 2021, 11:00 Uhr Ort: Videokonferenz (Zoom-Meeting, Informationen siehe unten) Referent: Stefan Kühnel, Dipl. Wirt.-Inf. Informatik 3 - Software Engineering Thema: Eine agile Methode zur simulativen Qualitätssicherung von Aktiven Sicherheitssystemen Kontext: Neben dem aktuellen Bestreben die Elektrizierung des Antriebs von Automobilen durch Innovationen voranzutreiben kommt der Integration von Fahrerassistenzsystemen im Rahmen der Automobilentwicklung eine besondere Bedeutung sowohl zur Steigerung des Fahrkomforts als auch zur Verbesserung der Sicherheit zu. Eine gewichtige Rolle spielen dabei Consumer Tests wie z.B. von Organisationen wie EuroNCAP, welche sowohl als Treiber aber auch zur Beurteilung der Leistungsfähigkeit der getesteten Sicherheitssysteme dient. Um die Qualität der dafür entwickelten Software besser beurteilen und steigern zu können, bilden Tests sowohl in ausgewiesenen Prüfarealen unter realen Bedingungen als auch simulationsbasierte Tests in synthetischen Umgebungen geeignete Ansätze, den Herausforderungen der kontinuierlichen Qualitätsverbesserung agil zu begegnen, wenngleich beide Ansätze über unterschiedliche Hürden und Grenzen verfügen. Speziell der hier betrachtete simulative Ansatz mündet nicht selten in dem Dilemma, dass die Entwicklung geeigneter Umgebungen ebenfalls ein hohes Maß an Ressourcen wie die eigentliche Systementwicklung benötigt und somit Letztere weiter parallel voranschreitet mit der Folge, dass die Simulationsumgebung aufgrund des hohen Aufwandes nicht rechtzeitig einsetzbar ist. Dies kann dann bspw. eintreten, wenn dem Aspekt der Modellbildung mit dem Güteziel einer möglichst engen Realitätsnähe ohne Rückkopplung auf die eingangs zu definierende Fragestellung: “Was soll durch die Simulation beantwortet werden?” erfolgt. Ziel: Das Ziel dieser Arbeit besteht darin, den komplexen Entwicklungsprozess von Aktiven Sicherheitssystemen im Rahmen der Consumer Tests durch einen simulativen Ansatz zur Verbesserung der Softwarequalität zu unterstützen. Weiterhin soll durch die Verhaltensanalyse von Algorithmen die Ressourcenallokation während der Entwicklung und der notwendigen realen Tests verbessert und damit effektiver gestaltet werden. Methode: Nach der Durchführung eines Systematic Literature Reviews (SLR) zu Prüfung evtl. vorhandener Ansätze und Methoden zur Entwicklung solcher Simulationsumgebungen wird eine eigene Methode entwickelt, vorgestellt und unter Berücksichtigung der vorliegenden Projektbedingungen diese Methode im Rahmen einer Fallstudie angewendet. Ergebnisse: Die Analyse des Projektkontextes kommt zu dem Ergebnis, dass es durchaus Simulationsaktivitäten gibt, jedoch eine strukturierte Herangehensweise zu ihrer Entwicklung fehlt. Das Systematic Literature Review bestätigt dieses Ergebnis, sodass der Bedarf der Entwicklung einer agilen Methode zur simulativen Qualitätssicherung von Aktiven Sicherheitssystemen insbesondere mit Blick auf Consumer Tests aufgezeigt wird. Die vorgestellte Methode umfasst vier Bausteine: (a) die Analyse und Modellierung des Untersuchungsraumes, (b) die Entwicklung einer adäquaten Simulationsinfrastruktur, (c) die Entwicklung von Bewertungsverfahren und (d) die Durchführung von Simulationsläufen sowie ihrer Auswertung. Zum Schluss wird die Methode mit Hilfe einer Fallstudie für ein Proof of Concept angewendet. Schlussfolgerung: Es wird aufgezeigt, dass die Methode während der qualitativen Bewertung von Softwarekomponenten auf simulativer Basis einen positiven Beitrag leistet, speziell dort, wo Äquivalenzklassentests nicht ausreichend sind, Consumer Test Szenarien hinreichend zu testen. Einschränkungen und Erweiterungsbedarf der Methode werden vorrangig in der Übertragung auf andere Kontexte im Fahrerassistenzumfeld und die Erweiterung um zusätzliche Consumer Test Szenarien wie den Fußgängerschutz gesehen. Es laden ein: die Dozentinnen und Dozenten der Informatik ******************************** Thema: [Promotion Stefan Kühnel] Vortrag Uhrzeit: 13. Nov. 2021 11:00 AM Amsterdam, Berlin, Rom, Stockholm, Wien Zoom-Meeting beitreten https://rwth.zoom.us/j/98203910423?pwd=UlRyWFNFSXpaNzZnQm9oZEE5T3lpZz09 Beste Grüße, Stefan Kühnel -- ---------------------------------------------------------------- Prof. Dr. Bernhard Rumpe | http://www.se-rwth.de Lehrstuhl Software Engineering | Informatik 3 Ahornstr. 55, 52074 Aachen, Germany | RWTH Aachen University Phone ++49 241 80-21301 / Fax -22218 |

1 0

Einladung: Informatik-Oberseminar Jens Deussen
by Jens Deussen 06 Dec '21

06 Dec '21

+********************************************************************** * * * Einladung * * * * Informatik-Oberseminar * * * +********************************************************************** Zeit: Montag, 13. Dezember 2021, 11.00 Uhr Ort: Zoom-Videokonferenz https://rwth.zoom.us/j/93783577046?pwd=NjFUNFZZUWpzNjhLN2VlOHEyVk9nQT09 Meeting-ID: 937 8357 7046 Kenncode: 561405 Referent: Jens Deussen M.Sc. Lehrstuhl Informatik 12: Software and Tools for Computational Engineering Thema: Global Derivatives Abstract: In this talk, we show how to obtain global derivatives that are guaranteed enclosures of the derivative information on specified domains. Therefore, we combine algorithmic differentiation (AD) methods with interval arithmetic and McCormick relaxations. While naive interval computations are prone to overestimation of exact value ranges, we identify special cases for which the natural interval extension applied to the AD methods compute exact value ranges for the global derivatives. We present two applications that benefit from global derivatives: deterministic global optimization by branch-and-bound methods, and significance-driven unreliable and approximate computing. Within the framework of the global optimization case study we introduce subdomain separability. This local property enables the partitioning of the optimization problem on subdomains that fulfill a certain monotonicity condition. The approximate computing case study demonstrates how to automatically prune artificial neural networks by using significance values. Es laden ein: die Dozentinnen und Dozenten der Informatik

1 0

Einladung: Informatik-Oberserminar Lucas Beyer
by Lucas Beyer 29 Nov '21

29 Nov '21

+********************************************************************** * * * Einladung * * * * Informatik-Oberseminar * * * +********************************************************************** Zeit: Freitag, 03. Dezember 2021, 10.00 Uhr Ort: Zoom Videokonferenz https://us02web.zoom.us/j/88210359250?pwd=QXFGMUEzTGRDVmJBbHgwT0lXcEhrdz09 Meeting-ID: 882 1035 9250 Kenncode: 542690 Referent: Lucas Beyer, Dipl.Ing Lehrstuhl Informatik 13 Thema: Deep Visual Human Sensing with Application in Robotics Abstract: In this talk, I present my thesis contributions to the field of visual human sensing that arise when deploying robots in environments with humans. After motivating the need for visual human sensing, we start by describing a novel human detector based on a 2D lidar sensor (e.g. a "laser scanner"). It is the first of its kind that is learning-based and general, specifically it does not encode a "two leg prior". Detection being covered, we move on to discuss person re-identification, and specifically our contribution of establishing triplet-loss based methods as a strong contender and principled approach in the field. Using this we also sketch the way to a completely novel approach on tracking which leverages such triplet-based re-identification models at its core. We then discuss more detailed analysis of individual persons, specifically their head orientation, which can serve as a cue for their intent or an indicator of what is interesting in the scene, among other things. We derive a novel cyclic regression loss based on the von-Mises distribution and use it, coupled with our "biternion" output layer, to learn continuous regression models using only discrete, weakly labeled data. Finally, we present a holistic system integrating all of these pieces and several more, highlighting the system-level difficulties of such integration, and proposing some ways around them. Es laden ein: die Dozentinnen und Dozenten der Informatik

1 0

Einladung Informatik Kolloquium Christopher Morris
by Grohe, Martin 24 Nov '21

24 Nov '21

*********************************************************************** * * * Einladung * * * * Informatik-Kolloquium * * * *********************************************************************** Zeit: Mittwoch, 1. Dezember 2021, 13.30 Uhr Ort: Zoom Videokonferenz https://rwth.zoom.us/j/95857189087?pwd=ajNJYUZFcHVvSHNFUmJya1RqUFhKUT09 Meeting-ID: 958 5718 9087 Kenncode: 050524 Referent: Christopher Morris, Quebec AI Institute and McGill University Thema: Learning with Graphs: From Theory to Applications Abstract: Graph-structured data is ubiquitous across domains ranging from chemo- and bioinformatics to image and social network analysis. To develop successful machine learning models in these domains, we need techniques mapping the graph's structure to a vectorial representation in a meaningful way---so-called graph embeddings. Starting from the 1960s in chemoinformatics, different research communities have worked in the area under various guises, often leading to recurring ideas. Moreover, triggered by the resurgence of (deep) neural networks, there is an ongoing trend in the machine learning community to design permutation-invariant or -equivariant neural architectures capable of dealing with graph input often denoted as neural graph networks (GNNs). However, although often successful in practice, GNN's capabilities and limits are understood to a lesser extend. In this talk, we overview some results shedding some light on the limitations and capabilities of GNNs by leveraging tools from graph theory and related areas. To complement the theory, we show how GNNs can act as an inductive bias to enhance state-of-art solvers for combinatorial optimization in a data-driven way. Es laden ein: die Dozentinnen und Dozenten der Informatik — Martin Grohe RWTH Aachen Lehrstuhl Informatik 7 Ahornstr. 55 52074 Aachen Germany e: grohe(a)informatik.rwth-aachen.de t: +49 241 80 21700 f: +49 241 8022 215

1 0

Einladung: Informatik-Oberserminar Theodora Kontogianni
by Theodora Kontogianni 24 Nov '21

24 Nov '21

+********************************************************************** * * * Einladung * * * * Informatik-Oberseminar * * * +********************************************************************** Zeit: Montag, 29. November 2021, 14.00 Uhr Ort: Zoom Videokonferenz https://rwth.zoom.us/j/93845227037?pwd=cm9qRjhtVm5JbWRYdGkrSUsyRythdz09 <https://www.google.com/url?q=https://rwth.zoom.us/j/93845227037?pwd%3Dcm9qR…> Meeting ID: 938 4522 7037 Passcode: 310833 Referent: Theodora Kontogianni, M.Sc. Lehrstuhl Informatik 13 Thema: Object Discovery, Interactive and 3D Segmentation for Large-Scale Computer Vision Tasks Abstract: In this talk, I present my thesis contributions that deal with issues arising when trying to exploit the large body of data available for computer vision tasks. In particular we address the problem of unsupervised object discovery in time-varying, large-scale image collections by proposing a novel tree structure that closely approximates the Minimum Spanning Tree and present an efficient construction approach along with an incremental update mechanism of the tree structure that incorporates new data as they are added to the image database. We then focus on defining novel 3D convolutional and recurrent operators over unstructured 3D point clouds. The goal is to learn point representations for the task of 3D semantic segmentation. We overcome the limitations of the unstructured and large-scale nature of the 3D point clouds by defining local structure through two clustering methods and expand the limited receptive field of previous approaches by modeling long-range relationships with the use of Recurrent Networks. In the third part, we address the task of interactive object segmentation where a computer vision algorithm segments an object aided by a human user. We present a method that significantly reduces the number of required user clicks compared to previous works. We use the sparse user corrections to adapt the model parameter on-the-fly during test time. In particular, we look at out-of-domain settings where the test datasets are significantly different from the datasets used to train our deep learning model. Es laden ein: die Dozentinnen und Dozenten der Informatik

1 0