November 2017 - ag-ullrich@ins.uni-bonn.de

[Ag-ullrich] [Ins-mitarbeiter] Seminarvortrag am Donnerstag, den 26.10. um 13:00 Uhr
by hosseini.ins.uni-bonn.de 13 Dec '17

13 Dec '17

Liebe Mitarbeiter des INS, im Rahmen unseres Forschungsseminars "Mathematics of Computation" wird am Donnerstag Albert Cohen (Université Pierre et Marie Curie, Paris) zum Thema 'Data assimilation, parameter estimation and reduced modeling' vortragen. Der Vortrag findet am Donnerstag, den 26.10. um 13:00 Uhr s.t. im Raum 6.020 WE6 statt. Weitere Informationen und eine Kurzzusammenfassung des Vortrags gibt es zeitnah unter http://ullrich.ins.uni-bonn.de/seminar/ . Mit den besten Grüßen Markus Bachmayr, Ira Neitzel, Tino Ullrich _______________________________________________ Ins-mitarbeiter mailing list Ins-mitarbeiter(a)ins.uni-bonn.de https://mail.ins.uni-bonn.de/mailman/listinfo/ins-mitarbeiter

1 11

[Ag-ullrich] [Ins-mitarbeiter] Einladung zum Promotionskolloquium am Freitag, 24.11. (10ct)
by Daniel Wissel 17 Nov '17

17 Nov '17

Liebe Kolleginnen und Kollegen, alle Interessierten möchte ich hiermit zu meinem Promotionskolloquium mit dem Thema "Schätzung der Intrinsischen Dimension mittels Simplex-Volumina" am nächsten Freitag, den 24.11. um 10:15 Uhr (10ct) in der Wegelerstr. 6, Raum 6.020 herzlich einladen. Im Anschluss wird es einen Umtrunk und einen kleinen Imbiss geben. Beste Grüße, Daniel -- Daniel Wissel Institut für Numerische Simulation, Universität Bonn Wegelerstr. 6, 53115 Bonn, Germany Office: We6 6.017 Phone: +49 228 733486 E-Mail: wissel(a)ins.uni-bonn.de _______________________________________________ Ins-mitarbeiter mailing list Ins-mitarbeiter(a)ins.uni-bonn.de https://mail.ins.uni-bonn.de/mailman/listinfo/ins-mitarbeiter

1 0

[Ag-ullrich] [Ins-mitarbeiter] BJF Vortraege
by Babette Weißkopf 16 Nov '17

16 Nov '17

Liebe INS-Mitarbeiterinnen und Mitarbeiter, von den ín der untenstehenden Vortragsankündigung der Bonn Junior Fellows ist der Vortrag von Herrn Feischel für das INS von besonderem Interesse. Daher bitte ich im Auftrag von Herrn Griebel um reges Erscheinen. *D**ienstag 28.11.2017* *14 Uhr Michael Feischl (Sydney): Adaptive algorithms magnetism and uncertainty* Vielen Dank. Mit besten Grüßen Bbaette Weißkopf -------- Forwarded Message -------- Subject: [Ins-prof] BJF Vortraege Date: Thu, 16 Nov 2017 10:05:33 +0100 From: Ira Neitzel <neitzel(a)ins.uni-bonn.de> To: ins-prof(a)ins.uni-bonn.de Liebe Kollegen, in der naechsten Woche wird es zwei BJF Vortraege im Numerikbereich geben: Montag 20.11.2017 12:00 Uhr Bernhard Schmitzer (Muenster): Optimal Transport: unbalanced extensions and efficient sparse algorithms (Lipschitz Saal) Freitag 24.11.2017 16 Uhr Ruming Zhang (Bremen): Scattering of waves by periodic structures (Lipschitz Saal) und noch einen in der folgenden Woche: *D**ienstag 28.11.2017* *14 Uhr Michael Feischl (Sydney): Adaptive algorithms magnetism and uncertainty* Am 20.11. und 24.11. wird es ein Abendessen mit den Kandidaten um 18:30 geben, am 28.11. ein Mittagessen um 12:15 Falls jemand mit einem der Kandidaten diskutieren moechte und am gemeinsame Essen teilnehmen, gebe ich das gern weiter. Viele Gruesse, Ira -- Prof. Dr. Ira Neitzel Institut für Numerische Simulation Universität Bonn Wegelerstr. 6 53115 Bonn Telefon: ++49 (0) 22873-3144 Internet:http://neitzel.ins.uni-bonn.de/ _______________________________________________ Ins-prof mailing list Ins-prof(a)ins.uni-bonn.de https://mail.ins.uni-bonn.de/mailman/listinfo/ins-prof _______________________________________________ Ins-mitarbeiter mailing list Ins-mitarbeiter(a)ins.uni-bonn.de https://mail.ins.uni-bonn.de/mailman/listinfo/ins-mitarbeiter

1 0

[Ag-ullrich] [Ins-mitarbeiter] Einladung zur Ada Lovelace-Preisverleihung
by Babette Weißkopf 16 Nov '17

16 Nov '17

Liebe Frau Prof. Neitzel, liebe INS-Professoren, liebe Mitarbeiterinnen und Mitarbeiter, auch in diesem Jahr verleiht das Institut für Numerische Simulation den Ada Lovelace-Preis für Mathematikerinnen. Die Kommission hat als Preisträgerin für das Akademische Jahr 2016/2017 Frau Clelia Albrecht für die Masterarbeit mit dem Titel "Parallelization of Adaptive Gradient Augmented Level Set Methods" als beste Nachwuchswissenschaftlerin am INS ausgewählt. Im Auftrag des Geschäftsführenden Direktors lade ich Sie hiermit herzlich zu der Preisverleihung am *_Donnerstag, den 14. Dezember 2017 um 12:00 Uhr c.t. ein. _ * Die Veranstaltung findet in der Wegelerstr. 6, 6. Etage, Raum 6.020 statt. Es wäre schön, wenn möglichst viele von Ihnen dabei wären, daher wird ausdrücklich darum gebeten, die Einladung wahrzunehmen. Im Anschluss an die Übergabe des Preises findet wie immer ein kleiner Umtrunk statt. Für eine bessere Planung bitte ich Sie, mir spätestens im Laufe der nächsten Woche mitzuteilen, ob Sie kommen werden. Vielen Dank. Mit freundlichen Grüßen Babette Weißkopf -- Babette Weißkopf Rheinische Friedrich-Wilhelms-Universität Bonn Institut für Numerische Simulation Wegelerstrasse 6 | 53115 Bonn E-Mail: weisskopf(a)ins.uni-bonn.de Tel.: ++49 228-733427 | Fax: ++49 228-737527 http://wissrech.ins.uni-bonn.de/main/ _______________________________________________ Ins-mitarbeiter mailing list Ins-mitarbeiter(a)ins.uni-bonn.de https://mail.ins.uni-bonn.de/mailman/listinfo/ins-mitarbeiter

1 0

[Ag-ullrich] [Ins-mitarbeiter] Fwd: PATC Course: Introduction to hybrid programming in HPC @ LRZ
by Babette Weisskopf 15 Nov '17

15 Nov '17

-------- Weitergeleitete Nachricht -------- Betreff: PATC Course: Introduction to hybrid programming in HPC @ LRZ Datum: Tue, 14 Nov 2017 11:52:01 +0100 (CET) Von: Rolf Rabenseifner <rabenseifner(a)hlrs.de> Antwort an: Rolf Rabenseifner <rabenseifner(a)hlrs.de>, Volker Weinberg <volker.weinberg(a)lrz.de> An: contact(a)ins.uni-bonn.de **************************************************************** ************** Registration deadline Jan 07, 2018 ************* ******************* Please register now ************************ Dear Madam or Sir / Sehr geehrte Damen und Herren, As hybrid programming techniques are getting more and more important in HPC, we provide a PRACE PATC Course: Introduction to hybrid programming in HPC LRZ, Garching near Munich Thursday, Jan 18, 2018, 10:00-17:00 Course Description: Most HPC systems are clusters of shared memory nodes. Such SMP nodes can be small multi-core CPUs up to large many-core CPUs. Parallel programming may combine the distributed memory parallelization on the node interconnect (e.g., with MPI) with the shared memory parallelization inside of each node (e.g., with OpenMP or MPI-3.0 shared memory). This course analyses the strengths and weaknesses of several parallel programming models on clusters of SMP nodes. Multi-socket- multi-core systems in highly parallel environments are given special consideration. MPI-3.0 has introduced a new shared memory programming interface, which can be combined with inter-node MPI communication. It can be used for direct neighbour accesses similar to OpenMP or for direct halo copies, and enables new hybrid programming models. These models are compared with various hybrid MPI+OpenMP approaches and pure MPI. Numerous case studies and micro-benchmarks demonstrate the performance-related aspects of hybrid programming. Tools for hybrid programming such as thread/process placement support and performance analysis are presented in a "how-to" section. This course provides scientific training in Computational Science, and in addition, the scientific exchange of the participants among themselves. Similar "MPI+X" tutorials about hybrid programming have been successfully presented by the lecturers Dr. Rolf Rabenseifner (HLRS, member of the steering committee of the MPI-3 Forum) and Dr. Georg Hager (RRZE, winner of the "Informatics Europe Curriculum Best Practices Award: Parallelism and Concurrency") during various supercomputing conferences in the past. Teachers: Dr. Georg Hager (RRZE/HPC Uni. Erlangen), Dr. Rolf Rabenseifner (HLRS) More information about the course and the registration form can be found online on the LRZ and PATC pages: https://www.lrz.de/services/compute/courses/2018-01-18_hhyp1w17/ and http://events.prace-ri.eu/e/LRZ-2018-HHYP1W17 Course: Jan 18, 2018, 10:00-17:00 Registration deadline: Jan 07, 2018 We would like to invite you to participate in this course at LRZ. Please also pass this workshop announcement to other interested colleagues. LRZ and HLRS are part of the Gauss Centre for Supercomputing (GCS), which is one of the six PRACE Advanced Training Centres (PATCs) that started in 2012. Kind regards, Rolf Rabenseifner (HLRS) and Volker Weinberg (LRZ) PS: Other upcoming courses: - Parallelization with MPI (Nov 20-22) Vienna [English] http://www.hlrs.de/training/2017/VSC3 - Shared memory parallelization with OpenMP (Nov 23-24) Vienna [Eng.] http://www.hlrs.de/training/2017/VSC4 - Advanced parallel programming with MPI and OpenMP (Nov 27-29) Juelich [Eng.] http://www.hlrs.de/training/2017/JSC - Fortran for Scientific Computing (Nov 27-Dec 01) Stuttgart [Ger.] http://www.hlrs.de/training/2017/FTN2 - Parallel Programming with MPI, OpenMP, and Tools (Feb 12-16) Dresden [Ger.] http://www.hlrs.de/training/2018/ZIH Information on further HPC courses: - by HLRS: http://www.hlrs.de/training/ and http://www.hlrs.de/training/overview/ - by LRZ: http://www.lrz.de/services/compute/courses/ - by the Gauss Centre of Supercomputing (GCS): http://www.gauss-centre.eu/training - by the PRACE Advanced Training Centres (PATCs): http://www.training.prace-ri.eu/ --------------------------------------------------------------------- HLRS Online Courses: http://www.hlrs.de/training/par-prog-ws/ As a member of the HLRS course mailing list, you should have received the password with an email titled "Get your access to HLRS online courses". --------------------------------------------------------------------- --------------------------------------------------------------------- We would appreciate if you could forward this email (without the following personal subscription-paragraph) to interested colleagues. --------------------------------------------------------------------- --------------------------------------------------------------------- Based on an upcoming law change, as mentioned in my prior email from Sep. 14, 2017, we kindly ask you to follow this link http://java.hlrs.de/subscription/elist?addr=contact@ins.uni-bonn.de&chk=208… to acknowledge your interest in continuing to receive mailings from HLRS. Please also note that if you do not acknowledge your agreement, it will no longer be possible for you to access our online course recordings. If you receive double postings or you want to stop my postings, then please unsubscribe at any time through visiting http://java.hlrs.de/subscription/elist?addr=contact@ins.uni-bonn.de&chk=208… --------------------------------------------------------------------- --------------------------------------------------------------------- Dr. Rolf Rabenseifner .. . . . . . . . . email rabenseifner(a)hlrs.de . High Performance Computing Center (HLRS) phone ++49(0)711/685-65530 . University of Stuttgart .. . . . . . . . fax : ++49(0)711/685-65832 . Head of Dpmt Parallel Computing ... www.hlrs.de/people/rabenseifner . Nobelstr. 19, D-70569 Stuttgart, Germany . . . (Office: Room 1.307) . --------------------------------------------------------------------- _______________________________________________ Ins-mitarbeiter mailing list Ins-mitarbeiter(a)ins.uni-bonn.de https://mail.ins.uni-bonn.de/mailman/listinfo/ins-mitarbeiter

1 0

[Ag-ullrich] [Ins-mitarbeiter] Fwd: Special Invitation: 6th Operations Research and Statistics Conference (ORS 2018)
by Babette Weisskopf 15 Nov '17

15 Nov '17

-------- Weitergeleitete Nachricht -------- Betreff: Special Invitation: 6th Operations Research and Statistics Conference (ORS 2018) Datum: Wed, 15 Nov 2017 17:43:51 +0800 Von: ORS Conference Secretariat <ors(a)cfp.globalstf.com.sg> Antwort an: secretariat(a)orstat.org Organisation: GSTF An: contact(a)ins.uni-bonn.de Dear Dr. Babette WeiBkopf, Hope this email finds you well. Given your interests and contributions to research and academia, we invite you/your research students to submit a paper to *6th Annual International Conference on Operations Research and Statistics (ORS 2018) *which will be held on 9^th - 10^th April 2018 in Singapore. The full paper submission deadline is on *15^th January 2018.* Hope that provides adequate time for you to complete the paper submission. If you need additional time,*_please email us at: _*secretariat(a)orstat.org <mailto:secretariat@orstat.org>_ _we may be able to extend additional time on a case by case basis. We hope you can be part of ORS 2018.** Briefly about ORS 2018: - *Keynote Speakers:* *Assoc. Prof. Ajay Jasra, Department of Statistics and Applied Probability, National University of Singapore * *Prof. C. Raju, *Professor of Quantitative Methods and Operations Management, Indian Institute of Management Kozhikode, India - *ORS 2018 Conference Proceedings: *Print ISSN: Print ISSN: 2251-1938, E-Periodical ISSN: 2251-1946 will be published and submitted to indexing partner organizations. - *Journal of Mathematics, Statistics and Operations Research: *All authors who present their papers at the conference will be invited to submit an extended version of their research paper for the GSTF Journal of Mathematics, Statistics and Operations Research (JMSOR) <http://lnk.ie/22ZYG/e=contact@ins.uni-bonn.de/http://globalstf.org/journal/…>- Print ISSN: 2251-3388, E-periodical: 2251-3396. All submitted papers will go through blind review process for acceptance and will be published without additional cost (digital). - *Best Paper Awards *and*Best Student Paper Awards *will be conferred at the conference (in order to qualify for the award, the paper must be presented at the conference). - *ORS 2018* will also constitute a *Special Panel Session*. - *Panel Proposals* are invited for submission. A minimum of three papers centering on a specific topic will be accepted for submission under *Panel Category* <http://lnk.ie/22ZYH/e=contact@ins.uni-bonn.de/http://orstat.org/SubmitPaper…>. For more information, please visit the *ORS 2018 *website: www.orstat.org <http://lnk.ie/22ZYI/e=contact@ins.uni-bonn.de/http://www.orstat.org/>.Would appreciate if you could disseminate this information to your colleagues/students who might be interested to participate in the conference. View testimonials from previous ORS conferences: Prof. Marcus Agustin <https://lnk.ie/22ZYJ/e=contact@ins.uni-bonn.de/https://www.youtube.com/watc…> Prof. Kourosh Marjani Rasmussen <https://lnk.ie/22ZYK/e=contact@ins.uni-bonn.de/https://www.youtube.com/watc…> Southern Illinois University Edwardsville, USA Technical University of Denmark Should you require any assistance or clarification, please do not hesitate to contact us at secretariat(a)orstat.org <mailto:secretariat@orstat.org>. Kind regards, ** *Moreganaraj* Programme Manager www.globalstf.org <http://lnk.ie/22ZYL/e=contact@ins.uni-bonn.de/http://www.globalstf.org> Global Science and Technology Forum (GSTF) 10 Anson Road, International Plaza, #13-12, Singapore- 079903 Phone: +65 6327 0166| Fax: +65 6327 0162 <https://lnk.ie/22ZYM/e=contact@ins.uni-bonn.de/https://lnk.ie/20I70/e=conta…><https://lnk.ie/22ZYN/e=contact@ins.uni-bonn.de/https://lnk.ie/20I71/e=conta…><file:///\\192.168.1.66\..\..\Training\AppData\Local\Microsoft\Windows\INetCache\Content.Outlook\244XNHPQ\youtube.com\globalstf> */The information in this e-mail is confidential and may be legally privileged. It is intended solely for the addressee(s). Access to this e-mail by anyone else is unauthorised./* */If you are not the intended recipient, any disclosure, copying, distribution or any action taken or omitted to be taken in reliance on it, is prohibited and may be unlawful./* If you would like to unsubscribe, please send ORS unsubscribe to unsubscribe(a)orstat.org <mailto:unsubscribe@orstat.org>. _______________________________________________ Ins-mitarbeiter mailing list Ins-mitarbeiter(a)ins.uni-bonn.de https://mail.ins.uni-bonn.de/mailman/listinfo/ins-mitarbeiter

1 0

[Ag-ullrich] [Ins-mitarbeiter] Fwd: Upcoming courses Nov 2017 - Feb 2018
by Babette Weisskopf 03 Nov '17

03 Nov '17

-------- Weitergeleitete Nachricht -------- Betreff: Upcoming courses Nov 2017 - Feb 2018 Datum: Thu, 2 Nov 2017 14:51:19 +0100 (CET) Von: Rolf Rabenseifner <rabenseifner(a)hlrs.de> Antwort an: Rolf Rabenseifner <rabenseifner(a)hlrs.de>, Lucienne Dettki <dettki(a)hlrs.de> An: contact(a)ins.uni-bonn.de Dear Madam or Sir / Sehr geehrte Damen und Herren, there are still free slots in our visualization course for some short term registrations. Here a list of upcoming courses with open registration: - Scientific Visualization (Nov 06-07) Stuttgart [English] http://www.hlrs.de/training/2017/VIS2 - Parallelization with MPI (Nov 20-22) Vienna [English] http://www.hlrs.de/training/2017/VSC3 - Shared memory parallelization with OpenMP (Nov 23-24) Vienna [Eng.] http://www.hlrs.de/training/2017/VSC4 - Fortran for Scientific Computing (Nov 27-Dec 01) Stuttgart [Ger.] http://www.hlrs.de/training/2017/FTN2 - Introduction to Hybrid Programming in HPC (Jan 18) Garching [Eng] http://www.hlrs.de/training/2018/HY-G - Parallel Programming with MPI, OpenMP, and Tools (Feb 12-16) Dresden [Ger.] http://www.hlrs.de/training/2018/ZIH Further courses, see http://www.hlrs.de/training/ Please, can you pass this course announcement also to interested colleagues. / Es waere schoen, wenn Sie diese Ankuendigung auch an interessierte Kollegen weitergeben koennten. Kind regards / Mit freundlichen Gruessen Rolf Rabenseifner & Lucienne Dettki --------------------------------------------------------------------- HLRS Online Courses: http://www.hlrs.de/training/par-prog-ws/ As a member of the HLRS course mailing list, you should have received the password with an email titled "Get your access to HLRS online courses". --------------------------------------------------------------------- --------------------------------------------------------------------- We would appreciate if you could forward this email (without the following personal subscription-paragraph) to interested colleagues. --------------------------------------------------------------------- --------------------------------------------------------------------- Based on an upcoming law change, as mentioned in my prior email from Sep. 14, 2017, we kindly ask you to follow this link http://java.hlrs.de/subscription/elist?addr=contact@ins.uni-bonn.de&chk=208… to acknowledge your interest in continuing to receive mailings from HLRS. Please also note that if you do not acknowledge your agreement, it will no longer be possible for you to access our online course recordings. If you receive double postings or you want to stop my postings, then please unsubscribe at any time through visiting http://java.hlrs.de/subscription/elist?addr=contact@ins.uni-bonn.de&chk=208… --------------------------------------------------------------------- --------------------------------------------------------------------- Dr. Rolf Rabenseifner .. . . . . . . . . . email rabenseifner(a)hlrs.de Lucienne Dettki . . . .. . . . . . . . . . . . . . . . dettki(a)hlrs.de High Performance Computing Center (HLRS) . phone ++49(0)711/685-65530 University of Stuttgart .. . . . . . . . . and : ++49(0)711/685-63894 Nobelstr. 19, D-70569 Stuttgart, Germany --------------------------------------------------------------------- _______________________________________________ Ins-mitarbeiter mailing list Ins-mitarbeiter(a)ins.uni-bonn.de https://mail.ins.uni-bonn.de/mailman/listinfo/ins-mitarbeiter

1 0

[Ag-ullrich] [Ins-mitarbeiter] Fwd: Fwd: Re: Hausdorff Forum, 03 November 2017, 14:15h
by Babette Weißkopf 03 Nov '17

03 Nov '17

-------- Weitergeleitete Nachricht -------- Betreff: Fwd: Re: Hausdorff Forum, 03 November 2017, 14:15h Datum: Fri, 3 Nov 2017 08:55:15 +0100 Von: Gunder Sievert <gunder-lily.sievert(a)hcm.uni-bonn.de> An: weisskopf(a)ins.uni-bonn.de Liebe Frau Weisskopf, bitte informieren Sie doch Ihre Mitarbeiter über u.s. Veranstaltung. Herzlichen Dank Gunder Sievert ______________________________________________ Gunder-Lily Sievert Hausdorff Center for Mathematics Rheinische Friedrich-Wilhelms-Universität Bonn Endenicher Allee 60, room 3.026 53115 Bonn Phone: +49(0)228 7362358 www.hcm.uni-bonn.de > English Version Below > > ------------------------------------------------------------------------------------------------------------------------------------------------ > > Sehr geehrte Damen und Herren, > > wir möchten Sie an das heutige Hausdorff Forum, im Mathematik-Zentrum > der Universität Bonn, Lipschitz Saal, Endenicher Allee 60 hinweisen. > > 14:15h Jürgen Gall, Universität Bonn: Analyzing Human Behavior in > Video Sequences > > Abstract: Analyzing the behavior of humans in continuous video > recordings requires to solve several tasks. This includes the > estimation and tracking of poses of multiple persons or the temporal > detection of activities. In this talk, I will describe some of our > recent works for this two tasks. The first task requires to solve > jointly the problem of person association over time and the pose > estimation for each person. This problem can be formulated as a graph > partitioning problem where a spatio-temporal graph is constructed from > detected body joints in a video. For the second task, temporal models > like recurrent neural networks are usually trained on videos that are > annotated at a frame-level. Acquiring such annotations, however, is > very time consuming and strong temporal models require large amounts > of annotated training data. Weaker forms of supervision like > transcripts are therefore investigated to learn temporal models. > > 15:15-15:45 Teepause > > > 15:15 h William A. P. Smith, University of York, UK: Model-based > analysis of faces > > Abstract: The quest to understand and model "face space" dates back to > the 1980s, though the variability and uniqueness of faces has > fascinated scholars, artists and scientists since antiquity. To learn > a face space from a sample of face data requires the factors that are > intrinsic to the face to be disentangled from extrinsic factors > related to the imaging environment. A model-based approach to analysis > of face images uses explicit models of the geometric and photometric > image formation processes in order to explain an image in terms of > factors such as shape, lighting and skin reflectance properties. This > is in contrast to learning-based approaches where a black box (usually > a convolutional neural network) is trained to directly classify or > regress some property of interest from an image. In this talk, I will > present a variety of work on model-based analysis of faces, describe > current work to try to integrate explicit models into black box > learning, present some applications including evaluation of > craniofacial surgical outcomes and discuss a collaboration with > psychologists to try to uncover the models and representations used in > human perception of faces. > > Mit freundlichen Grüßen > > Gunder-Lily Sievert > > ------------------------------------------------------------------------------------------------------------------------------------------------ > > > Dear Ladies and Gentlemen, > > > We are pleased to give you advance notice of today's Hausdorff Forum. > > Location: Lipschitz Hall, Endenicher Allee 60 > > > 14:15h Jürgen Gall, Universität Bonn: Analyzing Human Behavior in > Video Sequences > > Abstract: Analyzing the behavior of humans in continuous video > recordings requires to solve several tasks. This includes the > estimation and tracking of poses of multiple persons or the temporal > detection of activities. In this talk, I will describe some of our > recent works for this two tasks. The first task requires to solve > jointly the problem of person association over time and the pose > estimation for each person. This problem can be formulated as a graph > partitioning problem where a spatio-temporal graph is constructed from > detected body joints in a video. For the second task, temporal models > like recurrent neural networks are usually trained on videos that are > annotated at a frame-level. Acquiring such annotations, however, is > very time consuming and strong temporal models require large amounts > of annotated training data. Weaker forms of supervision like > transcripts are therefore investigated to learn temporal models. > > 15:15-15:45 Teepause > > 15:15 h William A. P. Smith, University of York, UK: Model-based > analysis of faces > > Abstract: The quest to understand and model "face space" dates back to > the 1980s, though the variability and uniqueness of faces has > fascinated scholars, artists and scientists since antiquity. To learn > a face space from a sample of face data requires the factors that are > intrinsic to the face to be disentangled from extrinsic factors > related to the imaging environment. A model-based approach to analysis > of face images uses explicit models of the geometric and photometric > image formation processes in order to explain an image in terms of > factors such as shape, lighting and skin reflectance properties. This > is in contrast to learning-based approaches where a black box (usually > a convolutional neural network) is trained to directly classify or > regress some property of interest from an image. In this talk, I will > present a variety of work on model-based analysis of faces, describe > current work to try to integrate explicit models into black box > learning, present some applications including evaluation of > craniofacial surgical outcomes and discuss a collaboration with > psychologists to try to uncover the models and representations used in > human perception of faces. > > > Best regards, > Gunder-Lily Sievert > -- > > ______________________________________________ > > Gunder-Lily Sievert > > Hausdorff Center for Mathematics > > Rheinische Friedrich-Wilhelms-Universität Bonn > > Endenicher Allee 60, room 3.026 > > 53115 Bonn > > Phone: +49(0)228 7362358 > > www.hcm.uni-bonn.de > > _______________________________________________ Ins-mitarbeiter mailing list Ins-mitarbeiter(a)ins.uni-bonn.de https://mail.ins.uni-bonn.de/mailman/listinfo/ins-mitarbeiter

1 0