Edwin Carlinet - HomePage

Phd. and Ing. in Image Processing

I received the Ing. diploma from Ecole Pour l'Informatique et les Techniques Avancées (EPITA) in 2011, specialized in Imaging and Scientific Computing. I then received the M.Sc. in Applied Mathematics for Computer Vision and Machine Learning with honors from Ecole Normale Supérieure de Cachan in 2012 and a Phd from Université Paris-Est in 2015.

I'm currently an assistant professor with the EPITA Research and Development Laboratory (LRDE)

My interests are twofold:

Image Processing and Statistical Learning. I have some affinities with Mathematical Morphology approaches to solve imaging problems.
Efficient programming and algorithmic optimization. I am an highly skilled C++ programmer interested in exploiting the new C++17 standards to get both efficient, readable and easy-to-write programs.

Teaching

EPITA

Parallel programming (materials)
Data compression (materials)
GPU Computing (materials)
Avdanced Algorithmics (TP 01 )

I am a teaching assistant for:

Théorie des Langages Rationnels (for 2nd year students) (16H)
Algorithmics (for 3rd year students) (24H)

Past

Sup'Biotech: junior lecturer for 2nd/3rd year students and in charge of the assistant team management.

. . In. , vol. () , pp. , , ed., . (paper) (slides) (poster)

(view online)

Research - Publications

Journal papers

@Article{	  carlinet.14.itip,
  author	= {Edwin Carlinet and Thierry G\'eraud},
  title		= {A Comparative Review of Component Tree Computation
		  Algorithms},
  journal	= {IEEE Transactions on Image Processing},
  year		= 2014,
  volume	= {23},
  number	= {9},
  month		= sep,
  pages		= {3885--3895},
  project	= {Image},
  abstract	= {Connected operators are morphological tools that have the
		  property of filtering images without creating new contours
		  and without moving the contours that are preserved. Those
		  operators are related to the max-tree and min-tree repre-
		  sentations of images, and many algorithms have been
		  proposed to compute those trees. However, no exhaustive
		  comparison of these algorithms has been proposed so far,
		  and the choice of an algorithm over another depends on many
		  parameters. Since the need for fast algorithms is obvious
		  for production code, we present an in-depth comparison of
		  the existing algorithms in a unique framework, as well as
		  variations of some of them that improve their efficiency.
		  This comparison involves both sequential and parallel
		  algorithms, and execution times are given with respect to
		  the number of threads, the input image size, and the pixel
		  value quantization. Eventually, a decision tree is given to
		  help the user choose the most appropriate algorithm with
		  respect to the user requirements. To favor reproducible
		  research, an online demo allows the user to upload an image
		  and bench the different algorithms, and the source code of
		  every algorithms has been made available.},
  urllrde	= {201406-ITIP},
  lrdeinc	= {Olena/MaxtreeReview},
  lrdekeywords	= {Image},
  lrdenewsdate	= {2014-06-16},
  lrdepaper	= {https://www.lrde.epita.fr/dload/papers/carlinet.14.itip.pdf}

}

@InProceedings{  carlinet.15.ipta,
	  author= {Edwin Carlinet and Thierry G\'eraud},
	  title= {Morphological Object Picking Based on the Color Tree of Shapes},
	  booktitle= {Proceedings of 5th International Conference on Image Processing Theory, Tools and Applications (IPTA'15)},
	  year= {2015},
	  address= {Orl{\'e}ans, France},
          month         = nov,
	  project= {Image},
	  urllrde= {201506-IPTA},
	  abstract= {The Tree of Shapes is a self-dual and contrast invariant
                  morphological tree that provides a high-level hierarchical
                  representation of images, suitable for many image
                  processing tasks. Despite its powerfulness and its
                  simplicity, it is still under-exploited in pattern
                  recognition and computer vision.  In this paper, we show
                  that both interactive and automatic image segmentation can
                  be achieved with some simple tree processings.  To that
                  aim, we rely on the ``Color Tree of Shapes'', recently
                  defined.  We propose a method for interactive segmentation
                  that does not involve any statistical learning, yet
                  yielding results that compete with state-of-the-art
                  approaches. We further extend this algorithm to
                  unsupervised segmentation and give some results.  Although
                  they are preliminary, they highlight the potential of such
                  an approach that works in the shape space.},
	  lrdepaper= {http://www.lrde.epita.fr/dload/papers/carlinet.15.ipta.pdf},
	  lrdekeywords= {Image},
	  lrdenewsdate= {2015-06-29},
          note          = {to appear},
}

@InProceedings{  carlinet.15.gretsi,
author= {Edwin Carlinet and Thierry G\'eraud},
title= {Une approche morphologique de segmentation interactive
                  avec l'arbre des formes couleur},
booktitle= {Actes du 25e Colloque du Groupe d'Etudes du Traitement du Signal et des Images (GRETSI'15)},
year= {2015},
address= {Lyon, France},
  month         = sep,
project= {Image},
urllrde= {201506-GRETSI},
abstract= {L'arbre des formes est un arbre morphologique \`a la fois
                  auto-dual et invariant par changement de contraste. Il
                  fournit une repr\'esentation haut-niveau de l'image,
                  int\'eressante pour de nombreuses t\^aches de traitement
                  d'images.  Malgr\'e son potentiel et sa simplicit\'e, il reste
                  largement sous-utilis\'e en reconnaissance des formes et
                  vision par ordinateur. Dans cet article, nous pr\'esentons
                  une m\'ethode de segmentation interactive qui s'effectue
                  simplement en manipulant cet arbre. Pour cela, nous nous
                  appuierons sur une repr\'esentation r\'ecemment d\'efinie~: \og
                  l'Arbre des Formes Couleur \fg. La m\'ethode de segmentation
                  interactive que nous proposons ne requiert aucun
                  apprentissage statistique~; n\'eanmoins elle obtient des
                  r\'esultats qui rivalisent avec ceux de l'\'etat de
                  l'art. Bien que pr\'eliminaires, les r\'esultats obtenus
                  mettent en avant le potentiel et l'int\'er\^et des m\'ethodes
                  travaillant dans l'espace des formes.},
lrdepaper= {http://www.lrde.epita.fr/dload/papers/carlinet.15.gretsi.pdf},
lrdekeywords= {Image},
lrdenewsdate= {2015-06-16},
  note          = {to appear},
}

@InProceedings{	  carlinet.15.ismm,
  author	= {Edwin Carlinet and Thierry G\'eraud},
  title		= {A Color Tree of Shapes with Illustrations on Filtering,
		  Simplification, and Segmentation},
  booktitle	= {Proceedings of the 12th International
		  Symposium on Mathematical Morphology (ISMM)},
  year		= {2015},
  series	= {Lecture Notes in Computer Science Series},
  volume	= {9082},
  address	= {Reykjavik, Iceland},
  publisher	= {Springer},
  editor	= {J.A. Benediktsson and J. Chanussot and L. Najman and
		   H. Talbot},
  pages		= {363--374},
  project	= {Image},
  urllrde	= {201503-ISMMa},
  abstract	= {The Tree of Shapes is a morphological tree that provides a
		  high-level, hierarchical, self-dual, and contrast invariant
		  representation of images, suitable for many image
		  processing tasks. When dealing with color images, one
		  cannot use the Tree of Shapes because its definition is
		  ill-formed on multivariate data. Common workarounds such as
		  marginal processing, or imposing a total order on data are
		  not satisfactory and yield many problems (color artifacts,
		  loss of invariances, etc.) In this paper, we highlight the
		  need for a self-dual and contrast invariant representation
		  of color images and we provide a method that builds a
		  single Tree of Shapes by merging the shapes computed
		  marginally, while guarantying the most important properties
		  of the ToS. This method does not try to impose an arbitrary
		  total ordering on values but uses only the inclusion
		  relationship between shapes. Eventually, we show the
		  relevance of our method and our structure through some
		  illustrations on filtering, image simplification, and
		  interactive segmentation.},
  lrdeslides	= {http://www.lrde.epita.fr/dload/papers/carlinet.15.ismm.slides.pdf},
  lrdepaper	= {http://www.lrde.epita.fr/dload/papers/carlinet.15.ismm.pdf},
  lrdekeywords	= {Image},
  lrdenewsdate	= {2015-04-07}
}

@InProceedings{	  geraud.15.ismm,
  author	= {Thierry G\'eraud and Edwin Carlinet and S\'ebastien Crozet},
  title		= {Self-Duality and Digital Topology: Links Between the
		  Morphological Tree of Shapes and Well-Composed Gray-Level
		  Images},
  booktitle	= {Proceedings of the 12th International
		  Symposium on Mathematical Morphology (ISMM)},
  year		= {2015},
  series	= {Lecture Notes in Computer Science Series},
  volume	= {9082},
  address	= {Reykjavik, Iceland},
  publisher	= {Springer},
  editor	= {J.A. Benediktsson and J. Chanussot and L. Najman and
		   H. Talbot},
  pages		= {573--584},
  project	= {Image},
  urllrde	= {201503-ISMMd},
  abstract	= {In digital topology, the use of a pair of connectivities
		  is required to avoid topological paradoxes. In mathematical
		  morphology, self-dual operators and methods also rely on
		  such a pair of connectivities. There are several major
		  issues: self-duality is impure, the image graph structure
		  depends on the image values, it impacts the way small
		  objects and texture are processed, and so on. A sub-class
		  of images defined on the cubical grid, {\it well-composed}
		  images, has been proposed, where all connectivities are
		  equivalent, thus avoiding many topological problems. In
		  this paper we unveil the link existing between the notion
		  of well-composed images and the morphological tree of
		  shapes. We prove that a well-composed image has a
		  well-defined tree of shapes. We also prove that the only
		  self-dual well-composed interpolation of a 2D image is
		  obtained by the median operator. What follows from our
		  results is that we can have a purely self-dual
		  representation of images, and consequently, purely
		  self-dual operators.},
  lrdepaper	= {http://www.lrde.epita.fr/dload/papers/geraud.15.ismm.pdf},
  lrdekeywords	= {Image},
  lrdenewsdate	= {2015-04-07}
}

@InProceedings{	  xu.15.ismm,
  author	= {Yongchao Xu and Edwin Carlinet and Thierry G\'eraud and
		  Laurent Najman},
  title		= {Efficient Computation of Attributes and Saliency Maps on
		  Tree-Based Image Representations},
  booktitle	= {Proceedings of the 12th International
		  Symposium on Mathematical Morphology (ISMM)},
  year		= {2015},
  series	= {Lecture Notes in Computer Science Series},
  volume	= {9082},
  address	= {Reykjavik, Iceland},
  publisher	= {Springer},
  editor	= {J.A. Benediktsson and J. Chanussot and L. Najman and
		   H. Talbot},
  pages		= {693--704},
  project	= {Image},
  urllrde	= {201503-ISMMc},
  abstract	= {Tree-based image representations are popular tools for
		  many applications in mathematical morphology and image
		  processing. Classically, one computes an attribute on each
		  node of a tree and decides whether to preserve or remove
		  some nodes upon the attribute function. This attribute
		  function plays a key role for the good performance of
		  tree-based applications. In this paper, we propose several
		  algorithms to compute efficiently some attribute
		  information. The first one is incremental computation of
		  information on region, contour, and context. Then we show
		  how to compute efficiently extremal information along the
		  contour (\eg, minimal gradient's magnitude along the
		  contour). Lastly, we depict computation of extinction-based
		  saliency map using tree-based image representations. The
		  computation complexity and the memory cost of these
		  algorithms are analyzed. To the best of our knowledge,
		  except information on region, none of the other algorithms
		  is presented explicitly in any state-of-the-art paper.},
  lrdepaper	= {http://www.lrde.epita.fr/dload/papers/xu.15.ismm.pdf},
  lrdekeywords	= {Image},
  lrdenewsdate	= {2015-04-07}
}

@InProceedings{	  carlinet.14.icip,
  author	= {Edwin Carlinet and Thierry G\'eraud},
  title		= {Getting a morphological Tree of Shapes for Multivariate
		  Images: Paths, Traps and Pitfalls},
  booktitle	= {Proceedings of the 21st International Conference on Image
		  Processing (ICIP)},
  year		= 2014,
  address	= {Paris, France},
  pages		= {615--619},
  project	= {Image},
  urllrde	= {201402-ICIPa},
  abstract	= {The Tree of Shapes is a morphological tree that provides
		  an high-level hierarchical representation of the image
		  suitable for many image processing tasks. This structure
		  has the desirable properties to be self-dual and
		  contrast-invariant and describes the organization of the
		  objects through level lines inclusion. Yet it is defined on
		  gray-level while many images have multivariate data (color
		  images, multispectral images\ldots) where information are
		  split across channels. In this paper, we propose some leads
		  to extend the tree of shapes on colors with classical
		  approaches based on total orders, more recent approaches
		  based on graphs and also a new distance-based method.
		  Eventually, we compare these approaches through denoising
		  to highlight their strengths and weaknesses and show the
		  strong potential of the new methods compared to classical
		  ones.},
  lrdeinc	= {Publications/carlinet.14.icip.inc},
  lrdepaper	= {http://www.lrde.epita.fr/dload/papers/carlinet.14.icip.pdf}
		  ,
  lrdeposter	= {http://www.lrde.epita.fr/dload/papers/carlinet.14.icip.poster.pdf}
		  ,
  lrdekeywords	= {Image},
  lrdenewsdate	= {2014-05-26}
}

@InProceedings{	  carlinet.14.icpr,
  author	= {Edwin Carlinet and Thierry G\'eraud},
  title		= {A Morphological Tree of Shapes for Color Images},
  booktitle	= {Proceedings of the 22nd International Conference on
		  Pattern Recognition (ICPR)},
  year		= 2014,
  month		= aug,
  address	= {Stockholm, Sweden},
  pages		= {1133--1137},
  project	= {Image},
  urllrde	= {201403-ICPR},
  lrdeinc	= {Publications/carlinet.14.icpr.inc},
  lrdekeywords	= {Image},
  lrdenewsdate	= {2014-04-02},
  lrdepaper	= {http://www.lrde.epita.fr/dload/papers/carlinet.14.icpr.pdf}
		  ,
  lrdeposter	= {http://www.lrde.epita.fr/dload/papers/carlinet.14.icpr.poster.pdf}

}

@InProceedings{	  xu.14.icip,
  author	= {Yongchao Xu and Edwin Carlinet and Thierry G\'eraud and
		  Laurent Najman},
  title		= {Meaningful disjoint level lines selection},
  booktitle	= {Proceedings of the 21st International Conference on Image
		  Processing (ICIP)},
  year		= 2014,
  address	= {Paris, France},
  pages		= {2938--2942},
  project	= {Image},
  urllrde	= {201402-ICIPe},
  lrdepaper	= {http://www.lrde.epita.fr/dload/papers/xu.14.icip.pdf},
  lrdeposter	= {http://www.lrde.epita.fr/dload/papers/xu.14.icip.poster.pdf}
		  ,
  abstract	= {Many methods based on the morphological notion of
		  \textit{shapes} (\textit{i.e.}, connected components of
		  level sets) have been proved to be very efficient in shape
		  recognition and shape analysis. The inclusion relationship
		  of the level lines (boundaries of level sets) forms the
		  tree of shapes, a tree-based image representation with a
		  high potential. Numerous applications using this tree
		  representation have been proposed. In this article, we
		  propose an efficient algorithm that extracts a set of
		  disjoint level lines in the image. These selected level
		  lines yields a simplified image with clean contours, which
		  also provides an intuitive idea about the main structure of
		  the tree of shapes. Besides, we obtain a saliency map
		  without transition problems around the contours by
		  weighting level lines with their significance. Experimental
		  results demonstrate the efficiency and usefulness of our
		  method. },
  lrdekeywords	= {Image},
  lrdenewsdate	= {2014-05-26}
}

@InProceedings{	  levillain.14.ciarp,
  author	= {Roland Levillain and Thierry G\'eraud and Laurent Najman
		  and Edwin Carlinet},
  title		= {Practical Genericity: Writing Image Processing Algorithms
		  Both Reusable and Efficient},
  booktitle	= {Proceedings of the 19th
		  Iberoamerican Congress on Pattern Recognition (CIARP)},
  address	= {Puerto Vallarta, Mexico},
  month		= nov,
  year		= {2014},
  pages		= {70--79},
  editor	= {Eduardo Bayro and Edwin Hancock},
  publisher	= {Springer-Verlag},
  series	= {Lecture Notes in Computer Science},
  volume	= {8827},
  project	= {Olena},
  urllrde	= {201409-CIARP},
  abstract	= {An important topic for the image processing and pattern
		  recognition community is the construction of open source
		  and efficient libraries. An increasing number of software
		  frameworks are said to be generic: they allow users to
		  write reusable algorithms compatible with many input image
		  types. However, this design choice is often made at the
		  expense of performance. We present an approach to preserve
		  efficiency in a generic image processing framework, by
		  leveraging data types features. Variants of generic
		  algorithms taking advantage of image types properties can
		  be defined, offering an adjustable trade-off between
		  genericity and efficiency. Our experiments show that these
		  generic optimizations can match dedicated code in terms of
		  execution times, and even sometimes perform better than
		  routines optimized by hand. Digital Topology software
		  should reflect the generality of the underlying
		  mathematics: mapping the latter to the former requires
		  genericity. By designing generic solutions, one can
		  effectively reuse digital topology data structures and
		  algorithms. We propose an image processing framework
		  focused on the Generic Programming paradigm in which an
		  algorithm on the paper can be turned into a single code,
		  written once and usable with various input types. This
		  approach enables users to design and implement new methods
		  at a lower cost, try cross-domain experiments and help
		  generalize results.},
  keywords	= {Generic Programming, Image Processing, Performance,
		  Olena},
  lrdepaper	= {http://www.lrde.epita.fr/dload/papers/levillain.14.ciarp.pdf}
		  ,
  lrdeslides	= {http://www.lrde.epita.fr/dload/papers/levillain.14.ciarp.slides.pdf}
		  ,
  lrdeprojects	= {Olena},
  lrdenewsdate	= {2014-09-10}
}

@InProceedings{	  carlinet.13.ismm,
  author	= {Edwin Carlinet and Thierry G\'eraud},
  title		= {A comparison of many max-tree computation algorithms},
  booktitle	= {Proceedings of the 11th International
		  Symposium on Mathematical Morphology (ISMM)},
  year		= 2013,
  editor	= {C.L. Luengo Hendriks and G. Borgefors and R. Strand},
  volume	= 7883,
  series	= {Lecture Notes in Computer Science Series},
  address	= {Uppsala, Sweden},
  publisher	= {Springer},
  pages		= {73--85},
  project	= {Image},
  urllrde	= {201305-ISMMd},
  abstract	= {With the development of connected filters in the last
		  decade, many algorithms have been proposed to compute the
		  max-tree. Max-tree allows computation of the most advanced
		  connected operators in a simple way. However, no exhaustive
		  comparison of these algorithms has been proposed so far and
		  the choice of an algorithm over another depends on many
		  parameters. Since the need for fast algorithms is obvious
		  for production code, we present an in depth comparison of
		  five algorithms and some variations of them in a unique
		  framework. Finally, a decision tree will be proposed to
		  help the user choose the most appropriate algorithm
		  according to their requirements.},
  lrdeposter	= {http://www.lrde.epita.fr/dload/papers/carlinet.13.ismm-poster.pdf}
		  ,
  lrdepaper	= {http://www.lrde.epita.fr/dload/papers/carlinet.13.ismm.pdf}
		  ,
  lrdekeywords	= {Image},
  lrdenewsdate	= {2013-03-14}
}

@InProceedings{	  geraud.13.ismm,
  author	= {Thierry G\'eraud and Edwin Carlinet and S\'ebastien Crozet
		  and Laurent Najman},
  title		= {A quasi-linear algorithm to compute the tree of shapes of
		  nD images},
  booktitle	= {Proceedings of the 11th International
		  Symposium on Mathematical Morphology (ISMM)},
  year		= 2013,
  editor	= {C.L. Luengo Hendriks and G. Borgefors and R. Strand},
  volume	= 7883,
  series	= {Lecture Notes in Computer Science Series},
  address	= {Uppsala, Sweden},
  publisher	= {Springer},
  pages		= {98--110},
  project	= {Image},
  urllrde	= {201305-ISMMa},
  abstract	= {To compute the morphological self-dual representation of
		  images, namely the tree of shapes, the state-of-the-art
		  algorithms do not have a satisfactory time complexity.
		  Furthermore the proposed algorithms are only effective for
		  2D images and they are far from being simple to implement.
		  That is really penalizing since a self-dual represen-
		  tation of images is a structure that gives rise to many
		  powerful operators and applications, and that could be very
		  useful for 3D images. In this paper we propose a
		  simple-to-write algorithm to compute the tree of shapes; it
		  works for nD images and has a quasi-linear complexity when
		  data quantization is low, typically 12 bits or less. To get
		  that result, this paper introduces a novel representation
		  of images that has some amazing properties of continuity,
		  while remaining discrete.},
  lrdepaper	= {http://www.lrde.epita.fr/dload/papers/geraud.2013.ismm.pdf}
		  ,
  lrdekeywords	= {Image},
  lrdenewsdate	= {2013-03-14}
}

Teaching

EPITA

Past

Research - Publications

Journal papers

Conference papers