BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Multi-Person Monocular 3D Pose Estimation Using Deep Neural Networ
 ks
DTSTART:20170627T103000
DTEND:20170627T123000
DTSTAMP:20260407T042019Z
UID:a0a4fe0cbe95bd4fe37892821f1372143a69f1055a50d9ac9345e33f
CATEGORIES:Conferences - Seminars
DESCRIPTION:Isinsu Katircioglu\nEDIC candidacy exam\nExam president: Prof.
  Mark Pauly\nThesis advisor: Prof. Pascal Fua\nThesis co-advisor: Dr. Math
 ieu Salzmann\nCo-examiner: Prof. Sabine Süsstrunk\n\nAbstract\nRecovering
  3D pose of multiple interacting subjects in unconstrained environments is
  highly challenging due to strong partial occlusions and ambiguities cause
 d by the close proximity of individuals. In this work\, we address the pro
 blem of multi- person 3D pose estimation from monocular RGB images using d
 eep neural networks. First we discuss three related methods that tackle di
 fferent aspects of our problem: a framework for 3D human pose prediction u
 sing shape context matching\, a deep learning model for multi-person 2D po
 se estimation that is cast as an integer linear programming problem and an
  approach for identifying collective activities from videos containing a g
 roup of interacting people. Finally\, we introduce our deep learning metho
 d that regresses individual 3D poses from heatmaps of 2D joint locations a
 nd feeds them to a dependency network to model the correlations between th
 em. We evaluate the performance of our approach on a new boxing dataset.\n
 Background papers\n\nDeeperCut: A Deeper\, Stronger\, and Faster Multi-Per
 son Pose Estimation Model. In ECCV'16\, by  Insafutdinov\, E.\, Pishchul
 in\, L.\, Andres\, B.\, Andriluka\, M.\, Schiele\, B.\nUnderstanding Co
 llective Activities of People from Videos. In PAMI'14\, by  Choi\, W.\, 
 Savarese S.\nRecovering 3D Human Body Configurations Using Shape Contexts.
  In PAMI'2006\, by Mori\, G.\, Malik\, J.\n\n 
LOCATION:BC 329 https://plan.epfl.ch/?room==BC%20329
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR