BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Holistic Sensing\, Estimating and Processing of Random Fields with
  Sensor Arrays.
DTSTART:20160718T093000
DTEND:20160718T113000
DTSTAMP:20260509T103426Z
UID:188f373dee58f95df54645027b2d7e160be764908ae7aa1f47a01ed9
CATEGORIES:Conferences - Seminars
DESCRIPTION:Matthieu Simeoni\nEDIC Candidacy Exam\nExam President: Prof. M
 ichael Unser\nThesis Director: Prof. Martin Vetterli\nThesis Co-director: 
 Prof. Victor Panaretos\nCo-examiner: Prof. Patrick Thiran\nBackground pape
 rs:The emerging field of signal processing on graphs: Extending high-dimen
 sional data analysis to networks and other irregular\ndomains\, by Shuman\
 , D. I.\, Narang\, S. K.\, Frossard\, P.\, Ortega\, A.\, & Vandergheynst\,
  P. Signal Processing Magazine IEEE\, 30(3)\, 83-98.Sparse sampling of sig
 nal innovations\, by Blu\, T.\, Dragotti\, P. L.\, Vetterli\, M.\, Marzili
 ano\, P.\, & Coulot\, L. Signal Processing Magazine\, IEEE\, 25(2)\, 31-40
 .Methodology and convergence rates for functional linear regression. The A
 nnals of Statistics\, 35(1)\, 70-91.Abstract\nMany scientific applications
  involve estimating the sufficient statistics of a physical phenomenon mod
 elled as a continuous random\nfield. In practice\, data is collected throu
 gh an acquisition system\, typically a sensor array\, consisting in a very
  large network of\nsensors\, filtering and sampling the incoming field at 
 different locations. Recently\, hierarchical designs have been proposed fo
 r these\narrays\, where groups of sensors are beamformed together so as to
  modify the properties of the spatial filtering performed by the\narray. M
 athematically speaking\, this acquisition system can be conveniently inter
 preted as a sampling operator\, chain of linear\noperators acting subseque
 ntly on the random field (filtering\, sampling\, beamforming\, etc). Formu
 lating the problem in such general\nterms permits to bring it into the sco
 pe of a variety of different methods\, such as Functional Data Analysis\, 
 Finite Rate of Innovation or\nGraph Signal Processing. In this thesis\, we
  adopt an holistic view on the system\, and propose inter-linked algorithm
 s for each of the\nindividual steps of the data processing pipeline\, incl
 uding\, but not restricted to: a versatile beamforming strategy to achieve
  a\nwide-range of spatial filters\; a resolution-free least-squares estima
 tion procedure based on functional linear regression\; and finally an\nalg
 orithm for extracting relevant features of the random field. For each of t
 he proposed algorithms and unlike state of the art methods\,\nwe work as m
 uch as possible at the analytical level\, considering the unknowns as cont
 inuous objects in some infinite-dimensional\nHilbert space. Discretisation
  is pushed to the very end of the processing chain\, for the sole purpose 
 of visualisation. We forecast that\nsuch an approach will lead to tremendo
 us improvements over state of the art methods in terms of accuracy\, numer
 ical stability\,\nmemory storage and computational resources.
LOCATION:BC 329 https://plan.epfl.ch/?room==BC%20329
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR