BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Physical Constrained Deep Neural Network
DTSTART:20210623T111500
DTEND:20210623T131500
DTSTAMP:20260407T181118Z
UID:93ecf03d6ab36aa34858a9263c4cd6fbf0e8fee60b5395aac6728251
CATEGORIES:Conferences - Seminars
DESCRIPTION:Zhen Wei\nEDIC candidacy exam\nexam president: Prof. Lenka Zde
 borova\nthesis advisor: Prof. Pascal Fua\nco-examiner: Prof. Jan Hesthaven
 \n\nAbstract\nShape optimization is an important technique that has been w
 idely applied in various engineering fields.\nHowever\, the high computati
 onal cost of simulation prohibits conventional\nshape optimization from be
 coming more efficient and effective.\nTo this end\, the problem of automat
 ic shape optimization is proposed to achieve a faster design process and\n
 significantly improved optimized shapes.\nIn this proposal\, the technical
  background for automatic shape optimization is discussed first.\nThen thr
 ee papers relating to this project are reviewed.\nSpecifically\, an attemp
 t that tackles the shape optimization problem with deep learning models is
 \nintroduced\, which serves as a basis for future studies in this project.
 \nTo solve the problems in the existing implementation and develop novel f
 unctionalities for shape optimization\,\nthe latest techniques of the mesh
  adaptation for CFD numerical methods and the graph learning models are re
 viewed\,\nproviding inspirations as well as potential solutions for future
  study. The major directions for further investigations are also\nelaborat
 ed in the last section of this proposal.\n\nBackground papers\n\n1. E. Rem
 elli et al.\, “MeshSDF: Differentiable Iso-Surface Extraction\,” in Ne
 urIPS\, 2020.\nhttps://papers.nips.cc/paper/2020/hash/fe40fb944ee700392ed5
 1bfe84dd4e3d-Abstract.html\n\n2. F. Xia et al.\, “Graph Learning: A Surv
 ey\,” IEEE Transactions on Artificial Intelligence\, 2021.\nhttps://ieee
 xplore.ieee.org/abstract/document/9416834\n\n3. F. Alauzet and A. Loseille
 \, “A decade of progress on anisotropic mesh adaptation for computationa
 l fluid dynamics\,” Computer-Aided Design\, 2016.\nhttps://www.sciencedi
 rect.com/science/article/pii/S0010448515001517\n 
LOCATION:
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR