BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:CESS seminar - Generating Semantic Building Information Over Time 
 and Space Using Visual Data
DTSTART:20230331T121500
DTEND:20230331T130000
DTSTAMP:20260410T031921Z
UID:732e499d44a80e75fa568ce36b5e3c8df3c65ae9ca7758b6f0d360f1
CATEGORIES:Conferences - Seminars
DESCRIPTION:Dr Iro Armeni\, Postdoctoral Researcher ETH-Zurich\nAbstract\n
 Many construction and operation processes\, such as those related to susta
 inable construction and circular economy\, require information on the stat
 e of the building (spatial)\, as well as on how it became what it is toda
 y (temporal). State-of-the-art approaches attempt to acquire spatial and t
 emporal building information using easily obtainable visual data and Compu
 ter Vision algorithms. However most of them examine a static version of o
 ur world when for many real-world applications we require a spatiotemporal
  interpretation of it. The examined change\, especially in the 3D domain\
 , is usually constrained to small spatial (e.g.\, that of a room) and temp
 oral (e.g.\, that of a few minutes or seconds) scales\, and mainly relates
  to relocation (movement of humans and objects). Considering the lifecycl
 e stages of the built environment\, from construction to operation and end
 -of-use\, consisting elements undergo changes that extend beyond relocatio
 n and relate to differences in the elements’ geometry\, appearance (e.g
 .\, floors before and after carpet is added)\, and topology (e.g.\, walls 
 begin as a group of studs and gradually change into the final wall structu
 re that users see). To address this\, I am employing the challenging setu
 p of construction sites (new or under renovation) as the grounds for devel
 oping computer vision methods that are able to work even in the most dras
 tic of changes. The first step is to register 3D point clouds of the same 
 space captured over time since they are almost never aligned in the same c
 oordinate system. In this talk I will present a benchmark and dataset for
  automatic partial and multi-way registration of the data regardless at wh
 ich point in time they were captured. This is a very challenging scenario 
 for computer vision algorithms due to the drastic changes in geometry\, a
 ppearance\, and topology that take place over time\, as well as the repeti
 tive structure of these scenes. I demonstrate the need for new algorithms
  that can handle such challenging scenes and drastic changes\, as well as 
 the relevance of the dataset. \n\n\nShort biography\nIro Armeni is a Post
 Doctoral Researcher at ETH Zurich and an incoming Assistant Professor at S
 tanford University (Sept. ’23)\, conducting interdisciplinary research b
 etween Architecture\, Civil Engineering\, and Visual Machine Perception. 
 Her area of focus is on automated semantic and operational understanding o
 f buildings throughout their life cycle using visual data. She completed 
 her PhD at Stanford University on August 2020\, Civil and Environmental En
 gineering Department\, with a PhD minor at the Computer Science Department
 . Prior to enrolling in the PhD program\, Iro received an MSc in Computer
  Science (Ionian University-2013)\, an MEng in Architecture and Digital De
 sign (University of Tokyo-2011)\, and a Diploma in Architectural Engineeri
 ng (National Technical University of Athens-2009). She is the recipient o
 f the ETHZ Postdoctoral Fellowship\, the Google PhD Fellowship on Machine 
 Perception\, and the Japanese Government (MEXT) scholarship. Iro has work
 ed as an architect and consultant for both the private and public sector.\
 n 
LOCATION:GC B1 10 https://plan.epfl.ch/?room==GC%20B1%2010 https://epfl.zo
 om.us/j/62283724904
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR