# Research

## Research.Main History

Changed lines 4-5 from:
Some work related to Kinect sensors.
to:
Some work related to Kinect sensors.
Changed lines 7-8 from:
A-contrario image segmentation (with code).
to:
A-contrario image segmentation (with code).
Changed line 10 from:
List of publications.
to:
List of publications.
Changed lines 4-5 from:
Some work related to Kinect sensors.
to:
Some work related to Kinect sensors.
Changed lines 7-8 from:
A-contrario image segmentation (with code).
to:
A-contrario image segmentation (with code).
Changed line 10 from:
List of publications.
to:
List of publications.
Changed lines 1-2 from:
(:title Research)
to:
(:title Research:)
Some work related to Kinect sensors.

* [[Research.Acsegmentor|Acsegmentor]]
A-contrario image segmentation (with code).

List of publications.
(:title Research)
Changed lines 1-2 from:
[[Research.Kinect|Kinect]]
[[Research.Publications|Publications]]
to:
* [[Research.Kinect|Kinect]]
* [[Research.Publications|Publications]]
Changed lines 1-2 from:
[[Research.Kinect] Kinect]
[[Research.Publications] Publications]
to:
[[Research.Kinect|Kinect]]
[[Research.Publications|Publications]]
Changed lines 1-2 from:
(:redirect [[Research.Publications]]:)
to:
[[Research.Kinect] Kinect]
[[Research.Publications] Publications]
Changed line 17 from:
!!!! [[#multitel09]] [2009] Monocular human upper body pose estimation for sign language analysis
to:
!!!! [[#multitel09]] [2009] Monocular human upper body pose estimation for sign language analysis %rfloat width=128px% Attach:signspeak_example.png
Changed lines 11-37 from:
!! [[#PhD]] PhD

!!!! Introduction

->
* '''Title''': ''"A contrario statistical learning and efficient vision systems to detect meaningful visual events"''
* '''Manuscript''': [[http://liris.cnrs.fr/Documents/Liris-3689.pdf | PDF (fr)]] \\

We aim at proposing robust and efficient algorithms to detect meaningful visual events.  Robustness implies, in particular, a close control of the number of false alarms made by an algorithm.  Since the a contrario statistical approach has proved to match this concern, e.g. to detect geometrical primitives, we extend it to applications where the existing purely analytical framework is not adapted. By combining analytical computations with Monte-Carlo simulations or statistical learning, we applied a contrario reasoning to problems such as image segmentation into homogeneous regions, which rely on multiple features and on data-driven exploration heuristics whose mathematical properties are difficult to determine.

To satisfy the speed requirement, we also study efficient architectures.  For low level vision, we experimented massive parallelism and developed a meaningful segments detection algorithm for programmable artificial retina, which operates in real-time.  For high level tasks, we propose an agent-based and parallel architecture combining information priorization, parallelism between processing levels and top-down / bottom-up communications to implement "anytime" algorithms which provide results all along their execution, the most salient first. This architecture is applied to object matching and shows promising results.

!!!! Segment extraction

These principles were first applied to '''segment extraction''' in images using the computational power of the digital retinas developed at ENSTA by [[http://www.ensta.fr/~tbernard|Thierry Bernard]], resulting into an efficient, massively parallel, statistically-founded and parameterless algorithm (see [[#acivs06 | acivs06]]).

!!!! A contrario image segmentation (acsegmentor)

The ''a contrario'' framework was then applied to a more complex problem, where exact computations are intractable: image '''segmentation into homogeneous regions'''. The resulting segmentation algorithm is called [[Research.Acsegmentor|'''acsegmentor''']] and can be used to filter out false alarms produced by existing algorithms. More details can be found on the [[Research.Acsegmentor | '''project page''']].

!!!! Object matching

Finally, we worked on '''object detection''' within a parallel architecture (logical, not physical this time). The goal is to detect objects stored in a database (one picture per object) in new images. Thanks to ''a contrario'' learning, several similarity measures can be used in a statistically founded framework to increase detection rates. Accurate ''a contrario'' distributions can be learned with as few as 10 natural images which do not contain the database objects. Combined with an adapted, agent-based architecture, we show that this approach is suitable for an "anytime" implantation (see [[#icvs08 | icvs08]]).

to:

!! [[#PhD]] PhD

!!!! Introduction

->
* '''Title''': ''"A contrario statistical learning and efficient vision systems to detect meaningful visual events"''
* '''Manuscript''': [[http://liris.cnrs.fr/Documents/Liris-3689.pdf | PDF (fr)]] \\

We aim at proposing robust and efficient algorithms to detect meaningful visual events.  Robustness implies, in particular, a close control of the number of false alarms made by an algorithm.  Since the a contrario statistical approach has proved to match this concern, e.g. to detect geometrical primitives, we extend it to applications where the existing purely analytical framework is not adapted. By combining analytical computations with Monte-Carlo simulations or statistical learning, we applied a contrario reasoning to problems such as image segmentation into homogeneous regions, which rely on multiple features and on data-driven exploration heuristics whose mathematical properties are difficult to determine.

To satisfy the speed requirement, we also study efficient architectures.  For low level vision, we experimented massive parallelism and developed a meaningful segments detection algorithm for programmable artificial retina, which operates in real-time.  For high level tasks, we propose an agent-based and parallel architecture combining information priorization, parallelism between processing levels and top-down / bottom-up communications to implement "anytime" algorithms which provide results all along their execution, the most salient first. This architecture is applied to object matching and shows promising results.

!!!! Segment extraction

These principles were first applied to '''segment extraction''' in images using the computational power of the digital retinas developed at ENSTA by [[http://www.ensta.fr/~tbernard|Thierry Bernard]], resulting into an efficient, massively parallel, statistically-founded and parameterless algorithm (see [[#acivs06 | acivs06]]).

!!!! A contrario image segmentation (acsegmentor)

The ''a contrario'' framework was then applied to a more complex problem, where exact computations are intractable: image '''segmentation into homogeneous regions'''. The resulting segmentation algorithm is called [[Research.Acsegmentor|'''acsegmentor''']] and can be used to filter out false alarms produced by existing algorithms. More details can be found on the [[Research.Acsegmentor | '''project page''']].

!!!! Object matching

Finally, we worked on '''object detection''' within a parallel architecture (logical, not physical this time). The goal is to detect objects stored in a database (one picture per object) in new images. Thanks to ''a contrario'' learning, several similarity measures can be used in a statistically founded framework to increase detection rates. Accurate ''a contrario'' distributions can be learned with as few as 10 natural images which do not contain the database objects. Combined with an adapted, agent-based architecture, we show that this approach is suitable for an "anytime" implantation (see [[#icvs08 | icvs08]]).
Changed line 49 from:
!!!! [[#pr09]] [2009] Image segmentation by a contrario simulation
to:
!!!! [[#pr09]] [2009] Image segmentation by a contrario simulation  %rfloat width=128px% Attach:segmentation_example.png
Changed lines 53-55 from:
* ''Abstract:'' %rfloat width=128px% Attach:segmentation_example.png
Segmenting an image into homogeneous regions generally involves a decision criterion to establish whether two adjacent regions are similar.  Decisions should be adaptive to get robust and accurate segmentation algorithms, avoid hazardous \textit{a priori} and have a clear interpretation.  We propose a decision process based on \textit{a contrario} reasoning: two regions are meaningfully different if the probability of observing such a difference in pure noise is very low.  Since the existing analytical methods are intractable in our case, we extend them to allow a mixed use of analytical computations and Monte-Carlo simulations.  The resulting decision criterion is tested experimentally through a simple merging algorithm, which can be used as a post-filtering and validation step for existing segmentation methods.
to:
* ''Abstract:'' Segmenting an image into homogeneous regions generally involves a decision criterion to establish whether two adjacent regions are similar.  Decisions should be adaptive to get robust and accurate segmentation algorithms, avoid hazardous \textit{a priori} and have a clear interpretation.  We propose a decision process based on \textit{a contrario} reasoning: two regions are meaningfully different if the probability of observing such a difference in pure noise is very low.  Since the existing analytical methods are intractable in our case, we extend them to allow a mixed use of analytical computations and Monte-Carlo simulations.  The resulting decision criterion is tested experimentally through a simple merging algorithm, which can be used as a post-filtering and validation step for existing segmentation methods.
Changed lines 53-55 from:
* ''Abstract:'' Segmenting an image into homogeneous regions generally involves a decision criterion to establish whether two adjacent regions are similar.  Decisions should be adaptive to get robust and accurate segmentation algorithms, avoid hazardous \textit{a priori} and have a clear interpretation.  We propose a decision process based on \textit{a contrario} reasoning: two regions are meaningfully different if the probability of observing such a difference in pure noise is very low.  Since the existing analytical methods are intractable in our case, we extend them to allow a mixed use of analytical computations and Monte-Carlo simulations.  The resulting decision criterion is tested experimentally through a simple merging algorithm, which can be used as a post-filtering and validation step for existing segmentation methods.
%rfloat width=128px% Attach:segmentation_example.png
to:
* ''Abstract:'' %rfloat width=128px% Attach:segmentation_example.png
Segmenting an image into homogeneous regions generally involves a decision criterion to establish whether two adjacent regions are
similar.  Decisions should be adaptive to get robust and accurate segmentation algorithms, avoid hazardous \textit{a priori} and have a clear interpretation.  We propose a decision process based on \textit{a contrario} reasoning: two regions are meaningfully different if the probability of observing such a difference in pure noise is very low.  Since the existing analytical methods are intractable in our case, we extend them to allow a mixed use of analytical computations and Monte-Carlo simulations.  The resulting decision criterion is tested experimentally through a simple merging algorithm, which can be used as a post-filtering and validation step for existing segmentation methods.
Changed lines 54-55 from:
%rfloat% %width=320px% Attach:segmentation_example.png
to:
%rfloat width=128px% Attach:segmentation_example.png
%rfloat% %width=320px% Attach:segmentation_example.png
October 27, 2010, at 08:34 PM by 163.117.150.243 -
Changed line 46 from:
* [[{$PubDirUrl}/files/research/burrus.09.multitel.monocular_body:pose.pdf | PDF (presentation)]] to: * [[{$PubDirUrl}/files/research/burrus.09.multitel.monocular_body_pose.pdf | PDF (presentation)]]
October 27, 2010, at 08:33 PM by 163.117.150.243 -
* [[{$PubDirUrl}/files/research/burrus.09.multitel.monocular_body:pose.pdf | PDF (presentation)]] October 07, 2009, at 05:00 PM by 163.117.150.243 - Changed line 123 from: * [[{$PubDirUrl}/files/research/burrus.05.report.segments_significatifs.pdf | PDF]]
to:
* [[{$PubDirUrl}/files/research/burrus.05.report.meaningful_segment_detection.pdf | PDF]] August 12, 2009, at 04:33 PM by 82.123.33.143 - Changed line 45 from: ''Talk in the 4th Multitel Spring Workshop, Mons, Belgium'' to: ''Talk at the 4th Multitel Spring Workshop, Mons, Belgium'' August 12, 2009, at 04:33 PM by 82.123.33.143 - Changed lines 46-47 from: to: * ''Abstract:'' We present a system to track human upper body using a single camera. The goal is to extract relevant features for sign language recognition, such as location, velocity and configuration of forearms and hands. Motion blur, rapid moves, self-occlusions and non-rigid deformations make the independent tracking of individual part difficult, ambiguous and not very reliable. Thus, we experiment top-down approaches based on pictorial models which aim at simultaneously modeling the geometry of human parts, the appearance of each part and the temporal continuity in a unified statistical framework. First results will be shown on the NGT corpus of Dutch sign language videos. August 12, 2009, at 04:32 PM by 82.123.33.143 - Added lines 43-46: !!!! [[#multitel09]] [2009] Monocular human upper body pose estimation for sign language analysis -> Nicolas Burrus and [[http://www.montefiore.ulg.ac.be/~piater/|Justus Piater]] ''Talk in the 4th Multitel Spring Workshop, Mons, Belgium'' August 12, 2009, at 04:29 PM by 82.123.33.143 - Changed line 43 from: !!!! [[#pr09]] [To appear] Image segmentation by a contrario simulation to: !!!! [[#pr09]] [2009] Image segmentation by a contrario simulation Deleted lines 47-53: !!!! [[#icvs08]] [2008] Bottom-up and top-down object matching using asynchronous agents and a contrario principles -> Nicolas Burrus and [[http://www.ensta.fr/~tbernard|Thierry M. Bernard]] and [[http://rfv.insa-lyon.fr/~jolion/|Jean-Michel Jolion]] ''In the Proceedings of the 6th International Conference on Computer Vision Systems (ICVS'08)'' * [[{$PubDirUrl}/files/research/burrus.08.icvs.object_matching.pdf | PDF]]
* [[https://liris.cnrs.fr/publis/?id=3445 | Link at Liris]]
* ''Abstract:'' We experiment a vision architecture for object matching based on a hierarchy of independent agents running asynchronously in parallel. Agents communicate through bidirectional signals, enabling the mix of top-down and bottom-up influences. Following the so-called a contrario principle, each signal is given a strength according to the statistical relevance of its associated visual data. By handling most important signals first, the system focuses on most promising hypotheses and provides relevant results as soon as possible. Compared to an equivalent feed-forward and sequential algorithm, our architecture is shown capable of handling more visual data and thus reach higher detection rates in less time.
@article{burrus2009pr,
title={{Image segmentation by a contrario simulation}},
author={Burrus, N. and Bernard, T.M. and Jolion, J.M.},
journal={Pattern Recognition},
volume={42},
number={7},
pages={1520--1532},
year={2009},
publisher={Elsevier}
}
@]

!!!! [[#icvs08]] [2008] Bottom-up and top-down object matching using asynchronous agents and a contrario principles
-> Nicolas Burrus and [[http://www.ensta.fr/~tbernard|Thierry M. Bernard]] and [[http://rfv.insa-lyon.fr/~jolion/|Jean-Michel Jolion]]
''In the Proceedings of the 6th International Conference on Computer Vision Systems (ICVS'08)''
* [[{$PubDirUrl}/files/research/burrus.08.icvs.object_matching.pdf | PDF]] * [[https://liris.cnrs.fr/publis/?id=3445 | Link at Liris]] * ''Abstract:'' We experiment a vision architecture for object matching based on a hierarchy of independent agents running asynchronously in parallel. Agents communicate through bidirectional signals, enabling the mix of top-down and bottom-up influences. Following the so-called a contrario principle, each signal is given a strength according to the statistical relevance of its associated visual data. By handling most important signals first, the system focuses on most promising hypotheses and provides relevant results as soon as possible. Compared to an equivalent feed-forward and sequential algorithm, our architecture is shown capable of handling more visual data and thus reach higher detection rates in less time. -> %box% [@ January 06, 2009, at 11:28 PM by 139.165.135.100 - Changed lines 18-21 from: * ''Title'': ''"A contrario statistical learning and efficient vision systems to detect meaningful visual events"'' * ''Manuscript'': [[http://liris.cnrs.fr/Documents/Liris-3689.pdf | PDF (fr)]] \\ to: * '''Title''': ''"A contrario statistical learning and efficient vision systems to detect meaningful visual events"'' * '''Manuscript''': [[http://liris.cnrs.fr/Documents/Liris-3689.pdf | PDF (fr)]] \\ January 06, 2009, at 11:28 PM by 139.165.135.100 - Added line 16: Changed lines 18-20 from: * Title: ''"A contrario statistical learning and efficient vision systems to detect meaningful visual events"'' * Manuscript: [[http://liris.cnrs.fr/Documents/Liris-3689.pdf | PDF (fr)]] \\ to: * ''Title'': ''"A contrario statistical learning and efficient vision systems to detect meaningful visual events"'' * ''Manuscript'': [[http://liris.cnrs.fr/Documents/Liris-3689.pdf | PDF (fr)]] \\ January 06, 2009, at 11:27 PM by 139.165.135.100 - Changed line 16 from: --> to: -> January 06, 2009, at 11:26 PM by 139.165.135.100 - Deleted line 17: January 06, 2009, at 11:26 PM by 139.165.135.100 - Added line 16: --> January 06, 2009, at 11:25 PM by 139.165.135.100 - Changed lines 16-19 from: Title: ''"A contrario statistical learning and efficient vision systems to detect meaningful visual events"'' '''Manuscript''': [[http://liris.cnrs.fr/Documents/Liris-3689.pdf | PDF (fr)]] \\ to: * Title: ''"A contrario statistical learning and efficient vision systems to detect meaningful visual events"'' * Manuscript: [[http://liris.cnrs.fr/Documents/Liris-3689.pdf | PDF (fr)]] \\ January 06, 2009, at 11:24 PM by 139.165.135.100 - Changed lines 18-19 from: ''Manuscrit'': [[http://liris.cnrs.fr/Documents/Liris-3689.pdf | PDF (fr)]] \\ to: '''Manuscript''': [[http://liris.cnrs.fr/Documents/Liris-3689.pdf | PDF (fr)]] \\ January 06, 2009, at 11:24 PM by 139.165.135.100 - Changed lines 18-19 from: Manuscrit: [[http://liris.cnrs.fr/Documents/Liris-3689.pdf | PDF (fr)]] \\ to: ''Manuscrit'': [[http://liris.cnrs.fr/Documents/Liris-3689.pdf | PDF (fr)]] \\ January 06, 2009, at 11:23 PM by 139.165.135.100 - Deleted lines 15-16: Original Subject [[{$PubDirUrl}/files/research/sujet-phd.pdf | PDF (fr)]] \\
January 06, 2009, at 11:22 PM by 139.165.135.100 -
Manuscrit: [[http://liris.cnrs.fr/Documents/Liris-3689.pdf | PDF (fr)]] \\
Changed line 44 from:
* [[{$PubDirUrl}/files/research/burrus.09.pr.segmentation.pdf | PDF]] to: * [[{$PubDirUrl}/files/research/burrus.09.pr.segmentation.pdf | PDF (preprint)]]
Changed lines 45-50 from:
* ''Abstract:''
Segmenting an image into homogeneous regions generally involves a decision criterion to establish whether two adjacent regions are similar.  Decisions should be adaptive to get robust and accurate segmentation algorithms, avoid
hazardous \textit{a priori} and have a clear interpretation.  We propose a decision process based on \textit{a contrario} reasoning: two regions are meaningfully different if the probability of observing such a difference in
pure noise is very low.  Since the existing analytical methods are intractable in our case, we extend them to allow a mixed use of analytical computations and Monte-Carlo simulations.  The resulting decision criterion is tested
experimentally through a simple merging algorithm, which can be used as a post-filtering and validation step for existing segmentation methods.
to:
* ''Abstract:'' Segmenting an image into homogeneous regions generally involves a decision criterion to establish whether two adjacent regions are similar.  Decisions should be adaptive to get robust and accurate segmentation algorithms, avoid hazardous \textit{a priori} and have a clear interpretation.  We propose a decision process based on \textit{a contrario} reasoning: two regions are meaningfully different if the probability of observing such a difference in pure noise is very low.  Since the existing analytical methods are intractable in our case, we extend them to allow a mixed use of analytical computations and Monte-Carlo simulations.  The resulting decision criterion is tested experimentally through a simple merging algorithm, which can be used as a post-filtering and validation step for existing segmentation methods.
Changed line 41 from:
!!!! [[#icvs08]] [2008] Bottom-up and top-down object matching using asynchronous agents and a contrario principles
to:
!!!! [[#pr09]] [To appear] Image segmentation by a contrario simulation
''Pattern Recognition journal''
* [[{$PubDirUrl}/files/research/burrus.09.pr.segmentation.pdf | PDF]] * ''Abstract:'' Segmenting an image into homogeneous regions generally involves a decision criterion to establish whether two adjacent regions are similar. Decisions should be adaptive to get robust and accurate segmentation algorithms, avoid hazardous \textit{a priori} and have a clear interpretation. We propose a decision process based on \textit{a contrario} reasoning: two regions are meaningfully different if the probability of observing such a difference in pure noise is very low. Since the existing analytical methods are intractable in our case, we extend them to allow a mixed use of analytical computations and Monte-Carlo simulations. The resulting decision criterion is tested experimentally through a simple merging algorithm, which can be used as a post-filtering and validation step for existing segmentation methods. !!!! [[#icvs08]] [2008] Bottom-up and top-down object matching using asynchronous agents and a contrario principles -> Nicolas Burrus and [[http://www.ensta.fr/~tbernard|Thierry M. Bernard]] and [[http://rfv.insa-lyon.fr/~jolion/|Jean-Michel Jolion]] Changed lines 34-35 from: Finally, we worked on '''object detection''' within a parallel architecture (logical, not physical this time). The goal is to detect objects stored in a database (one picture per object) in new images. Thanks to ''a contrario'' learning, several similarity measures can be used in a statistically founded framework to increase detection rates. Accurate a contrario distributions can be learned with as few as 10 natural images which do not contain the database objects. Combined with an adapted, agent-based architecture, we show that this approach is suitable for an "anytime" implantation. (see [[#icvs08 | icvs08]]) to: Finally, we worked on '''object detection''' within a parallel architecture (logical, not physical this time). The goal is to detect objects stored in a database (one picture per object) in new images. Thanks to ''a contrario'' learning, several similarity measures can be used in a statistically founded framework to increase detection rates. Accurate ''a contrario'' distributions can be learned with as few as 10 natural images which do not contain the database objects. Combined with an adapted, agent-based architecture, we show that this approach is suitable for an "anytime" implantation (see [[#icvs08 | icvs08]]). Changed lines 18-22 from: Title: ''"A contrario reasoning and efficient vision systems to detect meaningful visual events"'' We aim at proposing robust and efficient algorithms to detect meaningful visual events. Robustness implies, in particular, a close control of the number of false alarms made by an algorithm. Since the a contrario statistical approach has proved to match this concern, e.g. to detect geometrical primitives, we extend it to applications where the existing purely analytical framework is not adapted. By combining analytical computations with Monte-Carlo simulations or statistical learning, we applied a contrario reasoning to problems such as image segmentation into homogeneous regions, which rely on multiple features and on data-driven exploration heuristics whose mathematical properties are difficult to determine. to: Title: ''"A contrario statistical learning and efficient vision systems to detect meaningful visual events"'' We aim at proposing robust and efficient algorithms to detect meaningful visual events. Robustness implies, in particular, a close control of the number of false alarms made by an algorithm. Since the a contrario statistical approach has proved to match this concern, e.g. to detect geometrical primitives, we extend it to applications where the existing purely analytical framework is not adapted. By combining analytical computations with Monte-Carlo simulations or statistical learning, we applied a contrario reasoning to problems such as image segmentation into homogeneous regions, which rely on multiple features and on data-driven exploration heuristics whose mathematical properties are difficult to determine. Changed lines 30-35 from: The ''a contrario'' framework was then applied to a more complex problem, where exact computations are intractable: image '''segmentation into homogeneous regions'''. The resulting segmentation algorithm is called [[Research.Acsegmentor|'''acsegmentor''']]. More details can be found on the [[Research.Acsegmentor | '''project page''']]. !!!! Object matching (epav) I'm now working on '''object detection''' within a parallel architecture (logical, not physical this time). The goal is to detect objects stored in a database (one picture per object) in new images (see [[#icvs08 | icvs08]]) to: The ''a contrario'' framework was then applied to a more complex problem, where exact computations are intractable: image '''segmentation into homogeneous regions'''. The resulting segmentation algorithm is called [[Research.Acsegmentor|'''acsegmentor''']] and can be used to filter out false alarms produced by existing algorithms. More details can be found on the [[Research.Acsegmentor | '''project page''']]. !!!! Object matching Finally, we worked on '''object detection''' within a parallel architecture (logical, not physical this time). The goal is to detect objects stored in a database (one picture per object) in new images. Thanks to ''a contrario'' learning, several similarity measures can be used in a statistically founded framework to increase detection rates. Accurate a contrario distributions can be learned with as few as 10 natural images which do not contain the database objects. Combined with an adapted, agent-based architecture, we show that this approach is suitable for an "anytime" implantation. (see [[#icvs08 | icvs08]]) June 02, 2008, at 01:57 PM by 147.250.1.2 - Changed lines 50-59 from: title = {{Bottom-up and top-down object matching using asynchronous agents and a contrario principles}}, author = {Nicolas {Burrus} and Thierry {Bernard} and Jean-Michel {Jolion}}, year = {2008}, month = may, booktitle = {International Conference on Computer Vision Systems}, series = {Lecture Notes in Computer Science}, editor = {Antonios Gasteratos, Markus Vincze, John Tsotsos}, publisher = {Springer}, isbn = {978-3-540-79546-9}, url = {http://liris.cnrs.fr/publis/?id=3445}, to: title = {Bottom-Up and Top-Down Object Matching Using Asynchronous Agents and {\itshape a Contrario} Principles}, author = {Nicolas Burrus and Thierry M. Bernard and Jean-Michel Jolion}, editor = {Antonios Gasteratos and Markus Vincze and John K. Tsotsos}, booktitle = {Computer Vision Systems}, publisher = {Springer}, location = {Heidelberg}, series = {LNCS}, volume = {5008}, isbn = {978-3-540-79546-9}, pages = {343--352}, year = {2008}, url = {http://liris.cnrs.fr/publis/?id=3445}, May 07, 2008, at 02:05 PM by 147.250.1.2 - Changed line 58 from: isbn = {978-3-540-79546-}, to: isbn = {978-3-540-79546-9}, May 07, 2008, at 02:04 PM by 147.250.1.2 - Changed lines 35-36 from: I'm now working on '''object detection''' within a parallel architecture (logical, not physical this time). The goal is to detect objects stored in a database (one picture per object) in new images. to: I'm now working on '''object detection''' within a parallel architecture (logical, not physical this time). The goal is to detect objects stored in a database (one picture per object) in new images (see [[#icvs08 | icvs08]]) Changed line 42 from: !!!! [2008] Bottom-up and top-down object matching using asynchronous agents and a contrario principles to: !!!! [[#icvs08]] [2008] Bottom-up and top-down object matching using asynchronous agents and a contrario principles May 07, 2008, at 02:02 PM by 147.250.1.2 - Added line 17: Added line 19: May 07, 2008, at 01:59 PM by 147.250.1.2 - Changed line 40 from: !!!! [2008] Segmentation d'image par simulations a contrario ('''fr''') to: !!!! [2008] Bottom-up and top-down object matching using asynchronous agents and a contrario principles Changed lines 42-45 from: ''In the Proceedings of RFIA 2008'' * [[{$PubDirUrl}/files/research/burrus.08.rfia.segmentation.pdf | PDF]]
* [[https://liris.cnrs.fr/publis/?id=3321 | Link at Liris]]
* ''Abstract:'' Segmenting an image into homogeneous regions generally involves a decision criterion to establish whether two adjacent regions are similar. Decisions should be adaptive to get robust and accurate segmentation algorithms, avoid hazardous a priori and have clear justification. We propose a decision process based on an a contrario reasoning: two regions are meaningfully different if the probability of observing such a difference in pure noise is very low. Since the existing analytical methods are intractable in that case, we generalize them to allow a mixed use of analytical computations and Monte-Carlo simulations. The resulting decision criterion is tested experimentally through a simple merging algorithm, which can be used as a post-filtering and validation step for existing segmentation methods.
to:
''In the Proceedings of the 6th International Conference on Computer Vision Systems (ICVS'08)''
* [[{$PubDirUrl}/files/research/burrus.08.icvs.object_matching.pdf | PDF]] * [[https://liris.cnrs.fr/publis/?id=3445 | Link at Liris]] * ''Abstract:'' We experiment a vision architecture for object matching based on a hierarchy of independent agents running asynchronously in parallel. Agents communicate through bidirectional signals, enabling the mix of top-down and bottom-up influences. Following the so-called a contrario principle, each signal is given a strength according to the statistical relevance of its associated visual data. By handling most important signals first, the system focuses on most promising hypotheses and provides relevant results as soon as possible. Compared to an equivalent feed-forward and sequential algorithm, our architecture is shown capable of handling more visual data and thus reach higher detection rates in less time. Changed lines 47-48 from: @InProceedings{burrus08rfia, title = {{Segmentation d'image par simulations a contrario}}, to: @InProceedings{burrus08icvs, title = {{Bottom-up and top-down object matching using asynchronous agents and a contrario principles}}, Changed lines 51-54 from: month = jan, booktitle = {RFIA}, language = {fr}, url = {http://liris.cnrs.fr/publis/?id=3321}, to: month = may, booktitle = {International Conference on Computer Vision Systems}, series = {Lecture Notes in Computer Science}, editor = {Antonios Gasteratos, Markus Vincze, John Tsotsos}, publisher = {Springer}, isbn = {978-3-540-79546-}, url = {http://liris.cnrs.fr/publis/?id=3445 }, Added lines 61-78: !!!! [2008] Segmentation d'image par simulations a contrario ('''fr''') -> Nicolas Burrus and [[http://www.ensta.fr/~tbernard|Thierry M. Bernard]] and [[http://rfv.insa-lyon.fr/~jolion/|Jean-Michel Jolion]] ''In the Proceedings of RFIA 2008'' * [[{$PubDirUrl}/files/research/burrus.08.rfia.segmentation.pdf | PDF]]
* [[https://liris.cnrs.fr/publis/?id=3321 | Link at Liris]]
* ''Abstract:'' Segmenting an image into homogeneous regions generally involves a decision criterion to establish whether two adjacent regions are similar. Decisions should be adaptive to get robust and accurate segmentation algorithms, avoid hazardous a priori and have clear justification. We propose a decision process based on an a contrario reasoning: two regions are meaningfully different if the probability of observing such a difference in pure noise is very low. Since the existing analytical methods are intractable in that case, we generalize them to allow a mixed use of analytical computations and Monte-Carlo simulations. The resulting decision criterion is tested experimentally through a simple merging algorithm, which can be used as a post-filtering and validation step for existing segmentation methods.
-> %box% [@
@InProceedings{burrus08rfia,
title        = {{Segmentation d'image par simulations a contrario}},
author        = {Nicolas {Burrus} and Thierry {Bernard} and Jean-Michel {Jolion}},
year          = {2008},
month        = jan,
booktitle    = {RFIA},
language      = {fr},
url          = {http://liris.cnrs.fr/publis/?id=3321},
}
@]

May 07, 2008, at 01:56 PM by 147.250.1.2 -
Changed lines 21-23 from:
To satisfy the speed requirement, we also study efficient architectures.  For low level vision, we experimented massive
parallelism and developed a meaningful segments detection algorithm for programmable artificial retina, which operates in real-time.  For high level tasks, we propose an agent-based and parallel architecture combining information priorization, parallelism between processing levels and top-down / bottom-up communications to implement "anytime" algorithms which provide results all along their execution, the most salient first. This architecture is applied to object matching and shows promising results.
to:
To satisfy the speed requirement, we also study efficient architectures.  For low level vision, we experimented massive parallelism and developed a meaningful segments detection algorithm for programmable artificial retina, which operates in real-time.  For high level tasks, we propose an agent-based and parallel architecture combining information priorization, parallelism between processing levels and top-down / bottom-up communications to implement "anytime" algorithms which provide results all along their execution, the most salient first. This architecture is applied to object matching and shows promising results.
May 07, 2008, at 01:55 PM by 147.250.1.2 -
Changed lines 17-19 from:
Title: ''"Detection of meaningful visual events with fast information reduction operators."''
In two words: I try to find visual events (so far segments, homogeneous regions and objects) which are statistically meaningful and thus relevant for further analysis. I mainly rely on the ''a contrario'' framework initiated by the [[http://www.cmla.ens-cachan.fr/Cmla/index.html|CMLA]], which states that events are meaningful if their probability to occur in pure noise in very low. Such kind of statistical analysis requires many low-level computations, which can be done efficiently using massivelly parallel architectures.
to:
Title: ''"A contrario reasoning and efficient vision systems to detect meaningful visual events"''
We aim at proposing robust and efficient algorithms to detect meaningful visual events.  Robustness implies, in particular, a close
control of the number of false alarms made by an algorithm.  Since the a
contrario statistical approach has proved to match this concern, e.g. to detect geometrical primitives, we extend it to applications where the existing purely analytical framework is not adapted. By combining analytical computations with Monte-Carlo simulations or statistical learning, we applied a contrario reasoning to problems such as image segmentation into homogeneous regions, which rely on multiple features and on data-driven exploration heuristics whose mathematical properties are difficult to determine.

To satisfy the speed requirement, we also study efficient architectures.  For low level vision, we experimented massive
parallelism and developed a meaningful segments detection algorithm for programmable artificial retina, which operates in real-time.  For high level tasks, we propose an agent-based and parallel architecture combining information priorization, parallelism between processing levels and top-down / bottom-up communications to implement "anytime" algorithms which provide results all along their execution, the most salient first. This architecture is applied to object matching and shows promising results
.
Changed lines 32-35 from:
!!!! Object detection (epav)

I'm now working on '''object detection''' within a massivelly parallel architecture (logical, not physical this time). The goal is to detect objects stored in a database (one picture per object) in new images.
to:
!!!! Object matching (epav)

I'm now working on '''object detection''' within a parallel architecture (logical, not physical this time). The goal is to detect objects stored in a database (one picture per object) in new images.
Changed line 44 from:
@InProceedings{Liris-3321,
to:
@InProceedings{burrus08rfia,
Changed line 37 from:
!!!! [2008] Segmentation d'image par simulations a contrario ('''FRENCH''')
to:
!!!! [2008] Segmentation d'image par simulations a contrario ('''fr''')
Changed line 39 from:
''In the Proceedings of RFIA 2008'
to:
''In the Proceedings of RFIA 2008''
Changed lines 37-54 from:
to:
!!!! [2008] Segmentation d'image par simulations a contrario ('''FRENCH''')
-> Nicolas Burrus and [[http://www.ensta.fr/~tbernard|Thierry M. Bernard]] and [[http://rfv.insa-lyon.fr/~jolion/|Jean-Michel Jolion]]
''In the Proceedings of RFIA 2008'
* [[{$PubDirUrl}/files/research/burrus.08.rfia.segmentation.pdf | PDF]] * [[https://liris.cnrs.fr/publis/?id=3321 | Link at Liris]] * ''Abstract:'' Segmenting an image into homogeneous regions generally involves a decision criterion to establish whether two adjacent regions are similar. Decisions should be adaptive to get robust and accurate segmentation algorithms, avoid hazardous a priori and have clear justification. We propose a decision process based on an a contrario reasoning: two regions are meaningfully different if the probability of observing such a difference in pure noise is very low. Since the existing analytical methods are intractable in that case, we generalize them to allow a mixed use of analytical computations and Monte-Carlo simulations. The resulting decision criterion is tested experimentally through a simple merging algorithm, which can be used as a post-filtering and validation step for existing segmentation methods. -> %box% [@ @InProceedings{Liris-3321, title = {{Segmentation d'image par simulations a contrario}}, author = {Nicolas {Burrus} and Thierry {Bernard} and Jean-Michel {Jolion}}, year = {2008}, month = jan, booktitle = {RFIA}, language = {fr}, url = {http://liris.cnrs.fr/publis/?id=3321}, } @] Changed lines 61-62 from: * **Patent application** #20060229856 class: 703011000 (USPTO) to: * '''Patent application''' #20060229856 class: 703011000 (USPTO) Changed lines 61-62 from: * Patent application #20060229856 class: 703011000 (USPTO) to: * **Patent application** #20060229856 class: 703011000 (USPTO) Changed lines 24-25 from: !!!! Image segmentation (acsegmentor) to: !!!! A contrario image segmentation (acsegmentor) Changed lines 26-27 from: The ''a contrario'' framework was then applied to a more complex problem, where exact computations are intractable: image '''segmentation into homogeneous regions'''. The resulting segmentation algorithm is called [[Research.Acsegmentor|'''acsegmentor''']]. More details can be found on the [[Research.Acsegmentor | project page]]. to: The ''a contrario'' framework was then applied to a more complex problem, where exact computations are intractable: image '''segmentation into homogeneous regions'''. The resulting segmentation algorithm is called [[Research.Acsegmentor|'''acsegmentor''']]. More details can be found on the [[Research.Acsegmentor | '''project page''']]. Changed lines 26-27 from: The ''a contrario'' framework was then applied to a more complex problem, where exact computations are intractable: image '''segmentation into homogeneous regions'''. The resulting segmentation algorithm is called '''acsegmentor'''. More details can be found on the [[Research.Acsegmentor | project page]]. to: The ''a contrario'' framework was then applied to a more complex problem, where exact computations are intractable: image '''segmentation into homogeneous regions'''. The resulting segmentation algorithm is called [[Research.Acsegmentor|'''acsegmentor''']]. More details can be found on the [[Research.Acsegmentor | project page]]. Changed lines 30-31 from: I'm now working on '''object detection''' within a massivelly parallel architecture (logical, not hardware this time). The goal is to detect objects stored in a database (one picture per object) in new images. to: I'm now working on '''object detection''' within a massivelly parallel architecture (logical, not physical this time). The goal is to detect objects stored in a database (one picture per object) in new images. Changed lines 26-27 from: The ''a contrario'' framework was then applied to a more complex problem, where exact computations are intractable: image '''segmentation into homogeneous regions'''. The resulting segmentation algorithm is called '''acsegmentor'''. More details can be found on the [[Research.Acsegmentor|project page]]. to: The ''a contrario'' framework was then applied to a more complex problem, where exact computations are intractable: image '''segmentation into homogeneous regions'''. The resulting segmentation algorithm is called '''acsegmentor'''. More details can be found on the [[Research.Acsegmentor | project page]]. Changed lines 26-27 from: The ''a contrario'' framework was then applied to a more complex problem, where exact computations are intractable: image '''segmentation into homogeneous regions'''. The resulting segmentation algorithm is called '''acsegmentor'''. It can be downloaded on the [[Research.Acsegmentor|project page]]. Some publications are already submitted. to: The ''a contrario'' framework was then applied to a more complex problem, where exact computations are intractable: image '''segmentation into homogeneous regions'''. The resulting segmentation algorithm is called '''acsegmentor'''. More details can be found on the [[Research.Acsegmentor|project page]]. Added line 13: Deleted line 12: Changed lines 14-23 from: * Original Subject [[{$PubDirUrl}/files/research/sujet-phd.pdf | PDF (fr)]] \\
Title: ''"Detection of meaningful visual events with fast information reduction operators."''
In two words: I try to find visual events (so far segments, homogeneous regions and objects) which are statistically meaningful and thus relevant for further analysis. I mainly rely on the ''a contrario'' framework initiated by the [[http://www.cmla.ens-cachan.fr/Cmla/index.html|CMLA]], which states that events are meaningful if their probability to occur in pure noise in very low. Such kind of statistical analysis requires many low-level computations, which can be done efficiently using massivelly parallel architectures.

* These principles were first applied to '''segment extraction''' in images using the computational power of the digital retinas developed at ENSTA by [[http://www.ensta.fr/~tbernard|Thierry Bernard]], resulting into an efficient, massively parallel, statistically-founded and parameterless algorithm (see [[#acivs06 | acivs06]]).

* The ''a contrario'' framework was then applied to a more complex problem, where exact computations are intractable: image '''segmentation into homogeneous regions'''. The resulting segmentation algorithm is called '''acsegmentor'''. It can be downloaded on the [[Research.Acsegmentor|project page]]. Some publications are already submitted.

* I'm now working on '''object detection''' within a massivelly parallel architecture (logical, not hardware this time). The goal is to detect objects stored in a database (one picture per object) in new images.
to:
!!!! Introduction

Original Subject [[{$PubDirUrl}/files/research/sujet-phd.pdf | PDF (fr)]] \\ Title: ''"Detection of meaningful visual events with fast information reduction operators."'' In two words: I try to find visual events (so far segments, homogeneous regions and objects) which are statistically meaningful and thus relevant for further analysis. I mainly rely on the ''a contrario'' framework initiated by the [[http://www.cmla.ens-cachan.fr/Cmla/index.html|CMLA]], which states that events are meaningful if their probability to occur in pure noise in very low. Such kind of statistical analysis requires many low-level computations, which can be done efficiently using massivelly parallel architectures. !!!! Segment extraction These principles were first applied to '''segment extraction''' in images using the computational power of the digital retinas developed at ENSTA by [[http://www.ensta.fr/~tbernard|Thierry Bernard]], resulting into an efficient, massively parallel, statistically-founded and parameterless algorithm (see [[#acivs06 | acivs06]]). !!!! Image segmentation (acsegmentor) The ''a contrario'' framework was then applied to a more complex problem, where exact computations are intractable: image '''segmentation into homogeneous regions'''. The resulting segmentation algorithm is called '''acsegmentor'''. It can be downloaded on the [[Research.Acsegmentor|project page]]. Some publications are already submitted. !!!! Object detection (epav) I'm now working on '''object detection''' within a massivelly parallel architecture (logical, not hardware this time). The goal is to detect objects stored in a database (one picture per object) in new images. Changed lines 16-19 from: In two words: I try to find visual events (so far segments, homogeneous regions and objects) which are statistically meaningful and thus relevant for further analysis. I mainly rely on the ''a contrario'' framework initiated by the [[http://www.cmla.ens-cachan.fr/Cmla/index.html|CMLA]], which states that events are meaningful if their probability to occur in pure noise in very low. * These principles were first combined with the computational power of the digital retinas developed at ENSTA by [[http://www .ensta.fr/~tbernard|Thierry Bernard]] to make an efficient, massively parallel, statistically-founded and parameterless '''segment extraction algorithm''' (see [[#acivs06 | acivs06]]). to: In two words: I try to find visual events (so far segments, homogeneous regions and objects) which are statistically meaningful and thus relevant for further analysis. I mainly rely on the ''a contrario'' framework initiated by the [[http://www.cmla.ens-cachan.fr/Cmla/index.html|CMLA]], which states that events are meaningful if their probability to occur in pure noise in very low. Such kind of statistical analysis requires many low-level computations, which can be done efficiently using massivelly parallel architectures. * These principles were first applied to '''segment extraction ''' in images using the computational power of the digital retinas developed at ENSTA by [[http://www.ensta.fr/~tbernard|Thierry Bernard]], resulting into an efficient, massively parallel, statistically-founded and parameterless algorithm (see [[#acivs06 | acivs06]]). Changed lines 20-21 from: * The ''a contrario'' framework was then applied to a more complex problem, where exact computations are intractable: '''image segmentation into homogeneous regions'''. The resulting segmentation algorithm is called '''acsegmentor'''. It can be downloaded on the [[Research.Acsegmentor|project page]]. Some publications are already submitted. to: * The ''a contrario'' framework was then applied to a more complex problem, where exact computations are intractable: image '''segmentation into homogeneous regions'''. The resulting segmentation algorithm is called '''acsegmentor'''. It can be downloaded on the [[Research.Acsegmentor|project page]]. Some publications are already submitted. Changed lines 18-21 from: * These principles were first combined with the computational power of the digital retinas developed at ENSTA by [[http://www.ensta.fr/~tbernard|Thierry Bernard]] to make an efficient, statistically-founded and parameterless '''segment extraction algorithm''' (see [[#acivs06 | acivs06]]). * I then applied the ''a contrario'' framework to a more complex problem, where exact computations are intractable: image '''segmentation into homogeneous regions'''. The resulting segmentation algorithm is called [@acsegmentor@]. It can be downloaded on the [[Research.Acsegmentor|project page]]. Some publications are already submitted. to: * These principles were first combined with the computational power of the digital retinas developed at ENSTA by [[http://www.ensta.fr/~tbernard|Thierry Bernard]] to make an efficient, massively parallel, statistically-founded and parameterless '''segment extraction algorithm''' (see [[#acivs06 | acivs06]]). * The ''a contrario'' framework was then applied to a more complex problem, where exact computations are intractable: '''image segmentation into homogeneous regions'''. The resulting segmentation algorithm is called '''acsegmentor'''. It can be downloaded on the [[Research.Acsegmentor|project page]]. Some publications are already submitted. Changed lines 16-17 from: In two words: I try to find visual events (so far segments, homogeneous regions and objects) which are statistically meaningful and thus relevant for further analysis. I mainly rely on the ''a contrario'' framework initiated by the [[http://www.cmla.ens-cachan.fr/Cmla/index.html|CMLA]], which states that events are meaningful if there probability to occur in pure noise in very low. to: In two words: I try to find visual events (so far segments, homogeneous regions and objects) which are statistically meaningful and thus relevant for further analysis. I mainly rely on the ''a contrario'' framework initiated by the [[http://www.cmla.ens-cachan.fr/Cmla/index.html|CMLA]], which states that events are meaningful if their probability to occur in pure noise in very low. Added line 17: Changed lines 19-21 from: * I then applied the ''a contrario'' framework to a more complex problem, where exact computations are intractable: image segmentation into homogeneous regions. The resulting segmentation algorithm is called [@acsegmentor@]. It can be downloaded on the [[Research.Acsegmentor|project page]]. Some publications are submitted. * I'm now working on object detection within a massivelly parallel architecture (logical, not hardware this time). The goal is to detect objects stored in a database (one picture per object) in new images. to: * I then applied the ''a contrario'' framework to a more complex problem, where exact computations are intractable: image '''segmentation into homogeneous regions'''. The resulting segmentation algorithm is called [@acsegmentor@]. It can be downloaded on the [[Research.Acsegmentor|project page]]. Some publications are already submitted. * I'm now working on '''object detection''' within a massivelly parallel architecture (logical, not hardware this time). The goal is to detect objects stored in a database (one picture per object) in new images. Changed lines 16-18 from: * A contrario segments extraction (see [[#acivs06 | acivs06]]) * A contrario image segmentation: [[Research .Acsegmentor|Acsegmentor]] to: In two words: I try to find visual events (so far segments, homogeneous regions and objects) which are statistically meaningful and thus relevant for further analysis. I mainly rely on the ''a contrario'' framework initiated by the [[http://www.cmla.ens-cachan.fr/Cmla/index.html|CMLA]], which states that events are meaningful if there probability to occur in pure noise in very low. * These principles were first combined with the computational power of the digital retinas developed at ENSTA by [[http://www.ensta.fr/~tbernard|Thierry Bernard]] to make an efficient, statistically-founded and parameterless '''segment extraction algorithm''' (see [[#acivs06 | acivs06]]). * I then applied the ''a contrario'' framework to a more complex problem, where exact computations are intractable: image segmentation into homogeneous regions. The resulting segmentation algorithm is called [@acsegmentor@]. It can be downloaded on the [[Research.Acsegmentor|project page]]. Some publications are submitted. * I'm now working on object detection within a massivelly parallel architecture (logical, not hardware this time). The goal is to detect objects stored in a database (one picture per object) in new images. Changed lines 15-16 from: Titre: ''"Détection d'événements visuels saillants avec opérateurs rapides de réduction d'information"'' to: Title: ''"Detection of meaningful visual events with fast information reduction operators."'' * A contrario segments extraction (see [[#acivs06 | acivs06]]) Changed line 31 from: !!!! [2006] Adaptive Vision Leveraging Digital Retinas: Extracting Meaningful Segments to: !!!! [[#acivs06]] [2006] Adaptive Vision Leveraging Digital Retinas: Extracting Meaningful Segments Added line 35: * Slides of the oral presentation: [[{$PubDirUrl}/files/research/burrus.06.acivs.meaningful_segments_slides.pdf | PDF]]
Changed line 15 from:
Titre: ''Détection d'événements visuels saillants avec opérateurs rapides de réduction d'information''
to:
Titre: ''"Détection d'événements visuels saillants avec opérateurs rapides de réduction d'information"''
Changed lines 14-15 from:
* Original Subject [[{$PubDirUrl}/files/research/sujet-phd.pdf | PDF (fr)]] to: * Original Subject [[{$PubDirUrl}/files/research/sujet-phd.pdf | PDF (fr)]] \\
Titre: ''Détection d'événements visuels saillants avec opérateurs rapides de réduction d'information''
Changed line 14 from:
* Original Subject [[{$PubDirUrl/files/research/sujet-phd.pdf} | PDF (fr)]] to: * Original Subject [[{$PubDirUrl}/files/research/sujet-phd.pdf | PDF (fr)]]
Changed line 14 from:
* Subject
to:
* Original Subject [[{$PubDirUrl/files/research/sujet-phd.pdf} | PDF (fr)]] Changed line 14 from: * Subject of the PhD to: * Subject Added line 13: Added lines 11-15: !! [[#PhD]] PhD * Subject of the PhD * A contrario image segmentation: [[Research.Acsegmentor|Acsegmentor]] Changed lines 11-13 from: !! Publications to: !! [[#Publications]] Publications Changed lines 41-43 from: !!! C++ related to: !!! [[#CPP]] C++ related Changed lines 74-76 from: !!! Other to: !!! [[#Other]] Other Changed lines 14-16 from: !!! Computer Vision to: !!! [[#ComputerVision]] Computer Vision Deleted lines 6-8: !! Publications Added lines 11-13: !! Publications Changed lines 10-11 from: (:htoc Contents:) to: (:htoc:) Changed lines 12-13 from: \\ to: [[<<]] Changed lines 12-13 from: to: \\ Added line 13: Added line 12: Added lines 10-11: (:htoc Contents:) Deleted line 21: * {$PubDirUrl}/research/burrus.06.acivs.meaningful_segments.pdf
Changed line 43 from:
* [{{SERVER}}/files/research/burrus.04.report.metaprogramming.pdf PDF]
to:
* [[{$PubDirUrl}/files/research/burrus.04.report.metaprogramming.pdf | PDF]] Changed line 49 from: * [{{SERVER}}/files/research/burrus.03.mpool.scoop.pdf PDF] to: * [[{$PubDirUrl}/files/research/burrus.03.mpool.scoop.pdf | PDF]]
Changed lines 66-67 from:
* [{{SERVER}}/files/research/burrus.02.report.datatypes.pdf PDF]
to:
* [[{$PubDirUrl}/files/research/burrus.02.report.datatypes.pdf | PDF]] * [[http://www.lrde.epita.fr/cgi-bin/twiki/view/Publications/20020925-Seminar-Burrus-Report | Link at LRDE]] Changed lines 76-77 from: * [{{SERVER}}/files/research/burrus_lesage.03.report.evidence.pdf PDF] * [http://www.lrde.epita.fr/cgi-bin/twiki/view/Publications/20030528-Seminar-TheoryOfEvidence Link at LRDE] to: * [[{$PubDirUrl}/files/research/burrus_lesage.03.report.evidence.pdf | PDF]]
* [[http://www.lrde.epita.fr/cgi-bin/twiki/view/Publications/20030528-Seminar-TheoryOfEvidence | Link at LRDE]]
Changed line 83 from:
* [{{SERVER}}/files/research/csi2004.01.report.nn.pdf PDF]
to:
* [[{$PubDirUrl}/files/research/csi2004.01.report.nn.pdf | PDF]] Changed lines 22-23 from: * {$FileDir}/research/burrus.06.acivs.meaningful_segments.pdf
* [[{$FileDir}/research/burrus.06.acivs.meaningful_segments.pdf | PDF]] to: * {$PubDirUrl}/research/burrus.06.acivs.meaningful_segments.pdf
* [[{$PubDirUrl}/files/research/burrus.06.acivs.meaningful_segments.pdf | PDF]] Changed line 29 from: * [[{$FileDir}/research/burrus.05.report.segments_significatifs.pdf | PDF]]
to:
* [[{$PubDirUrl}/files/research/burrus.05.report.segments_significatifs.pdf | PDF]] Changed lines 22-23 from: * [[research/burrus.06.acivs.meaningful_segments.pdf | PDF]] to: * {$FileDir}/research/burrus.06.acivs.meaningful_segments.pdf
* [[{$FileDir}/ research/burrus.06.acivs.meaningful_segments.pdf | PDF]] Changed line 22 from: * [[{$FileDir}/research/burrus.06.acivs.meaningful_segments.pdf | PDF]]
to:
* [[research/burrus.06.acivs.meaningful_segments.pdf | PDF]]
Changed line 22 from:
* [[{$FILES}/research/burrus.06.acivs.meaningful_segments.pdf | PDF]] to: * [[{$FileDir}/research/burrus.06.acivs.meaningful_segments.pdf | PDF]]
Changed line 28 from:
* [[{$FILES}/research/burrus.05.report.segments_significatifs.pdf | PDF]] to: * [[{$FileDir}/research/burrus.05.report.segments_significatifs.pdf | PDF]]
Changed line 16 from:
* [http://www.iti.tugraz.at/dsc06/CR/dsc06_p03_cr.pdf PDF]
to:
* [[http://www.iti.tugraz.at/dsc06/CR/dsc06_p03_cr.pdf | PDF]]
Changed line 22 from:
* [{$FILES}/burrus.06.acivs.meaningful_segments.pdf PDF] to: * [[{$FILES}/research/burrus.06.acivs.meaningful_segments.pdf | PDF]]
Changed line 28 from:
* [{{SERVER}}/files/research/burrus.05.report.segments_significatifs.pdf PDF]
to:
* [[{$FILES}/research/burrus.05.report.segments_significatifs.pdf | PDF]] Changed line 22 from: * [Path:../pub/files/burrus.06.acivs.meaningful_segments.pdf PDF] to: * [{$FILES}/burrus.06.acivs.meaningful_segments.pdf PDF]
Changed line 22 from:
* [Path:pub/files/burrus.06.acivs.meaningful_segments.pdf PDF]
to:
* [Path:../pub/files/burrus.06.acivs.meaningful_segments.pdf PDF]
Changed line 22 from:
* [burrus.06.acivs.meaningful_segments.pdf PDF]
to:
* [Path:pub/files/burrus.06.acivs.meaningful_segments.pdf PDF]
Changed lines 3-4 from:
%define=box block border="1px solid #bbbbbb"%
to:
%define=box block border="0px solid #bbbbbb"%
Changed lines 3-4 from:
%define=box block border="1px solid #bbbbbbbb"%
to:
%define=box block border="1px solid #bbbbbb"%
Changed lines 3-4 from:
%define=box block border="0px solid #aaaaaa"%
to:
%define=box block border="1px solid #bbbbbbbb"%
Changed lines 3-4 from:
%define=box block border="1px solid #aaaaaa"%
to:
%define=box block border="0px solid #aaaaaa"%
Changed line 52 from:
-> %box% @[
to:
-> %box% [@
Changed lines 52-53 from:
%box%\

[=
to:
-> %box% @[
Changed lines 61-62 from:
=]
to:
@]
Changed line 53 from:
[@
to:
[=
Changed lines 62-63 from:
@]
to:
=]
Changed lines 52-53 from:
%box%[@
to:
%box%\

[@
Changed lines 51-53 from:
* ''Abstract:'' Object-oriented and generic programming are both supported in C++. OOP provides high expressiveness whereas GP leads to more efficient programs by avoiding dynamic typing. This paper presents SCOOP, a new paradigm which enables both classical OO design and high performance in C++ by mixing OOP and GP. We show how classical and advanced OO features such as virtual methods, multiple inheritance, argument covariance, virtual types and multimethods can be implemented in a fully statically typed model, hence without run-time overhead.
%box%
[@
to:
* ''Abstract:'' Object-oriented and generic programming are both supported in C++. OOP provides high expressiveness whereas GP leads to more efficient programs by avoiding dynamic typing. This paper presents SCOOP, a new paradigm which enables both classical OO design and high performance in C++ by mixing OOP and GP. We show how classical and advanced OO features such as virtual methods, multiple inheritance, argument covariance, virtual types and multimethods can be implemented in a fully statically typed model, hence without run-time overhead.
%box%[@
Changed lines 52-53 from:
%box%[@
to:
%box%

[@
%define=box block border="1px solid #aaaaaa"%
Changed line 52 from:
%apply=pre border=1px solid #AAAAAA%[@
to:
%box%[@
Changed line 50 from:
(:markup:)[@
to:
%apply=pre border=1px solid #AAAAAA%[@
Changed line 50 from:
[@
to:
(:markup:)[@
Changed line 50 from:
[@
to:
[@
Changed lines 59-60 from:
@]
to:
@]
Changed lines 50-51 from:
<blockquote>
<pre>
to:
[@
Changed lines 59-61 from:
</pre>
</blockquote>

to:
@]
Changed lines 33-34 from:
:* Patent application #20060229856 class: 703011000 (USPTO)
to:
* Patent application #20060229856 class: 703011000 (USPTO)
Changed line 20 from:
* [{{SERVER}}/files/research/burrus.06.acivs.meaningful_segments.pdf PDF]
to:
* [burrus.06.acivs.meaningful_segments.pdf PDF]
Changed lines 23-32 from:
==== '''[2005] Détection de segments significatifs sur rétine artificielle programmable (fr)''' ====
:
Nicolas Burrus
: ''Rapport de stage de master''
:*
[{{SERVER}}/files/research/burrus.05.report.segments_significatifs.pdf PDF]
:* ''Résumé:'' Ce travail se place à l’intersection de deux approches à priori indépendantes du traitement des images : une approche matérielle et algorithmique basée sur l’architecture cellulaire massivement parallèle des rétines ar ti&#64257;cielles ; et une approche a contrario statistiquement fondée de la perception cherchant à minimiser le nombre de paramètres nécessaires pour analyser les images. Nous nous proposons d’étudier la combinaison de ces deux univers à travers un opérateur simple et générique : l’extraction de segments signi&#64257;catifs.

==== '''[2004] Visualization and White Matter Fiber Tracking with Diffusion Tensor Magnetic Resonance Images''' ====
:
Nicolas Burrus
: ''Internship report, Siemens Corporate Research (Princeton)''
:* ''Abstract:'' We propose to model DT-MRI fiber tracking using particle filters. The resulting algorithm can handle both noise and ambiguities raised by partial volume effects and crossing fibers. Combined with a simple clustering algorithm, distinct fiber branches can be explored without loss of tracking power.  Interesting results were obtained with a reasonable computation time.  Last, the flexibility of the framework makes it possible to add many parameters to the model, opening good perspectives for future improvements.
to:
!!!! [2005] Détection de segments significatifs sur rétine artificielle programmable (fr)
-> Nicolas Burrus
''Rapport de stage de master''
*
[{{SERVER}}/files/research/burrus.05.report.segments_significatifs.pdf PDF]
* ''Résumé:'' Ce travail se place à l’intersection de deux approches à priori indépendantes du traitement des images : une approche matérielle et algorithmique basée sur l’architecture cellulaire massivement parallèle des rétines ar ti&#64257;cielles ; et une approche a contrario statistiquement fondée de la perception cherchant à minimiser le nombre de paramètres nécessaires pour analyser les images. Nous nous proposons d’étudier la combinaison de ces deux univers à travers un opérateur simple et générique : l’extraction de segments signi&#64257;catifs.

!!!! [2004] Visualization and White Matter Fiber Tracking with Diffusion Tensor Magnetic Resonance Images
-> Nicolas Burrus
''Internship report, Siemens Corporate Research (Princeton)''
* ''Abstract:'' We propose to model DT-MRI fiber tracking using particle filters. The resulting algorithm can handle both noise and ambiguities raised by partial volume effects and crossing fibers. Combined with a simple clustering algorithm, distinct fiber branches can be explored without loss of tracking power.  Interesting results were obtained with a reasonable computation time.  Last, the flexibility of the framework makes it possible to add many parameters to the model, opening good perspectives for future improvements.
Changed lines 35-48 from:
=== C++ related ===

==== '''[2004]
Introduction to C++ metaprogramming''' ====
:
Nicolas Burrus
: ''Tutorial''
:* [{{SERVER}}/files/research/burrus.04.report.metaprogramming.pdf PDF]
:* ''Abstract:'' This report aims at simplifying the discovery of the static C++ world. Mostly relying on Todd Veldhuisen Techniques for scientific C++ and Andrei Alexandrescu Modern C++ Design, we try to give to the novice metaprogrammer most of the basics notions he should learn, in a didactic way. The goal is also to make the reader work and think by himself before discovering already existing solutions, in order to facilitate the understanding.

==== '''[2003] A Static C++ Object-Oriented Programming (SCOOP) Paradigm Mixing Benefits of Traditional OOP and Generic Programming''' ====
:
Nicolas Burrus, Alexandre Duret-Lutz, Thierry Geraud, David Lesage and Raphaël Poss
: ''In the Proceedings of the Workshop on Multiple Paradigm with OO Languages (MPOOL'03) Anaheim, CA, USA Oct. 2003''
:* [{{SERVER}}/files/research/burrus.03.mpool.scoop.pdf PDF]
:* [http://www
:* ''Abstract:'' Object-oriented and generic programming are both supported in C++. OOP provides high expressiveness whereas GP leads to more efficient programs by avoiding dynamic typing. This paper presents SCOOP, a new paradigm which enables both classical OO design and high performance in C++ by mixing OOP and GP. We show how classical and advanced OO features such as virtual methods, multiple inheritance, argument covariance, virtual types and multimethods can be implemented in a fully statically typed model, hence without run-time overhead.
to:
!!! C++ related

!!!! [2004]
Introduction to C++ metaprogramming
-> Nicolas Burrus
''Tutorial''
* [{{SERVER}}/files/research/burrus.04.report.metaprogramming.pdf PDF]
* ''Abstract:'' This report aims at simplifying the discovery of the static C++ world. Mostly relying on Todd Veldhuisen Techniques for scientific C++ and Andrei Alexandrescu Modern C++ Design, we try to give to the novice metaprogrammer most of the basics notions he should learn, in a didactic way. The goal is also to make the reader work and think by himself before discovering already existing solutions, in order to facilitate the understanding.

!!!! [2003] A Static C++ Object-Oriented Programming (SCOOP) Paradigm Mixing Benefits of Traditional OOP and Generic Programming
-> Nicolas Burrus, Alexandre Duret-Lutz, [[http://www.lrde.epita.fr/~theo|Thierry Geraud]], [[http://hogur.org|David Lesage]] and Raphaël Poss
''In the Proceedings of the Workshop on Multiple Paradigm with OO Languages (MPOOL
'03) Anaheim, CA, USA Oct. 2003''
* [{{SERVER}}/files/research/burrus.03.mpool.scoop.pdf PDF
]
* [[http://www.lrde.epita.fr/cgi-bin/twiki/view/Publications/200310-MPOOL | Link at LRDE]]

* ''Abstract:'' Object-oriented and generic programming are both supported in C++. OOP provides high expressiveness whereas GP leads to more efficient programs by avoiding dynamic typing. This paper presents SCOOP, a new paradigm which enables both classical OO design and high performance in C++ by mixing OOP and GP. We show how classical and advanced OO features such as virtual methods, multiple inheritance, argument covariance, virtual types and multimethods can be implemented in a fully statically typed model, hence without run-time overhead.
Changed lines 63-83 from:
==== '''[2002] Safe and efficient data types in C++ (Intègre)''' ====
:
Nicolas Burrus
: ''LRDE technical report''
:*
[{{SERVER}}/files/research/burrus.02.report.datatypes.pdf PDF]
:* ''Abstract:'' Using C++ builtin types is very unsafe as they are inherited from C types, which do not have overflow checking and have dangerous side effects and unexpected behaviors. Using intensive meta programming, it becomes possible to design safe data types with a minimal runtime overhead. As we want to be able to use existing algorithms, these types have to interact transparently with C++ builtin types. Primarily designed for Olena, a generic image processing library, our work needs to provide mechanisms to allow easy integration in generic algorithms.

=== Other ===

==== '''[2003]
Theory of Evidence''' ====
:
Nicolas Burrus and David Lesage
: ''LRDE technical report''
:*
[{{SERVER}}/files/research/burrus_lesage.03.report.evidence.pdf PDF]
:* ''Abstract:'' The theory of evidence, also called Dempster-Shafer theory or belief functions theory, has been introduced by Glenn Shafer in 1976 as a new approach for representing uncertainty. Nowadays, this formalism is considered as one of the most interesting alternatives to Bayesian networks and fuzzy sets. This report makes an overview of both theoretical and implementation aspects of this theory. After a short survey of the historical motivations for this theory, we present its interesting properties through the Transferable Belief Model formalism. From a more practical point of view, we propose a review of the existing optimizations for facing the #P complexity of Dempster-Shafer computations. This report introduces a new, promising concept to compute repeated fusions: the delayed mass valuation. Finally, we present Evidenz, our general-purpose C++ library for designing efficient Dempster-Shafer engines.

==== '''[2001] Neural Networks: Multi-Layer Perceptron and Hop&#64257;eld Network''' ====
:
Sylvain Berlemont, Nicolas Burrus, David Lesage, Francis Maes, Jean-Baptiste Mouret, Benoît Perrot, Maxime Rey, Nicolas Tisserand, Astrid Wang
: ''LRDE technical report''
:*
[{{SERVER}}/files/research/csi2004.01.report.nn.pdf PDF]
:* ''Abstract:'' The ability of neural networks to derive meaning from complicated or imprecise data make them a powerful tool to extract hidden correlations between patterns or to recognize noised patterns. This report is dedicated to the study of Multi-Layer Perceptrons (MLP) and Hop&#64257;eld networks. In particular, two applications are detailed. MLP possibilities are illustrated through an image compression software and Hop&#64257;eld networks are studied through a character recognizer. For both applications, theoretical principles, heuristic and algorithmic improvements are discussed thanks to various experiments.
to:
!!!! [2002] Safe and efficient data types in C++ (Intègre)
-> Nicolas Burrus
''LRDE technical report''
*
[{{SERVER}}/files/research/burrus.02.report.datatypes.pdf PDF]
* ''Abstract:'' Using C++ builtin types is very unsafe as they are inherited from C types, which do not have overflow checking and have dangerous side effects and unexpected behaviors. Using intensive meta programming, it becomes possible to design safe data types with a minimal runtime overhead. As we want to be able to use existing algorithms, these types have to interact transparently with C++ builtin types. Primarily designed for Olena, a generic image processing library, our work needs to provide mechanisms to allow easy integration in generic algorithms.

!!! Other

!!!! [2003]
Theory of Evidence
-> Nicolas Burrus and David Lesage
''LRDE technical report''
*
[{{SERVER}}/files/research/burrus_lesage.03.report.evidence.pdf PDF]
* ''Abstract:'' The theory of evidence, also called Dempster-Shafer theory or belief functions theory, has been introduced by Glenn Shafer in 1976 as a new approach for representing uncertainty. Nowadays, this formalism is considered as one of the most interesting alternatives to Bayesian networks and fuzzy sets. This report makes an overview of both theoretical and implementation aspects of this theory. After a short survey of the historical motivations for this theory, we present its interesting properties through the Transferable Belief Model formalism. From a more practical point of view, we propose a review of the existing optimizations for facing the #P complexity of Dempster-Shafer computations. This report introduces a new, promising concept to compute repeated fusions: the delayed mass valuation. Finally, we present Evidenz, our general-purpose C++ library for designing efficient Dempster-Shafer engines.

!!!! [2001] Neural Networks: Multi-Layer Perceptron and Hop&#64257;eld Network
-> Sylvain Berlemont, Nicolas Burrus, David Lesage, Francis Maes, Jean-Baptiste Mouret, Benoît Perrot, Maxime Rey, Nicolas Tisserand, Astrid Wang
''LRDE technical report''
*
[{{SERVER}}/files/research/csi2004.01.report.nn.pdf PDF]
* ''Abstract:'' The ability of neural networks to derive meaning from complicated or imprecise data make them a powerful tool to extract hidden correlations between patterns or to recognize noised patterns. This report is dedicated to the study of Multi-Layer Perceptrons (MLP) and Hop&#64257;eld networks. In particular, two applications are detailed. MLP possibilities are illustrated through an image compression software and Hop&#64257;eld networks are studied through a character recognizer. For both applications, theoretical principles, heuristic and algorithmic improvements are discussed thanks to various experiments.
Changed lines 17-22 from:
==== '''[2006] Adaptive Vision Leveraging Digital Retinas: Extracting Meaningful Segments''' ====
:
Nicolas Burrus and Thierry Bernard
: ''In the Proceedings of Advanced Concepts for Intelligent Vision Systems International Conference (ACIVS'06), 2006''
:* [{{SERVER}}/files/research/burrus.06.acivs.meaningful_segments.pdf PDF]
:* ''Abstract:'' In general, the less probable an event, the more attention we pay to it. Likewise, considering visual perception, it is interesting to regard important image features as those that most depart from randomness. This statistical approach has recently led to the development of adaptive and parameterless algorithms for image analysis. However, they require computer-intensive statistical measurements. Digital retinas, with their massively parallel and collective computing capababilities, seem adapted to such computational tasks. These principles and opportunities are investigated here through a case study: extracting meaningful segments from an image.
to:
!!!! [2006] Adaptive Vision Leveraging Digital Retinas: Extracting Meaningful Segments
-> Nicolas Burrus and [[http://www.ensta.fr/~tbernard|Thierry Bernard]]
''In the Proceedings of Advanced Concepts for Intelligent Vision Systems International Conference
(ACIVS'06), 2006''
* [{{SERVER}}/files/research/burrus.06.acivs.meaningful_segments.pdf PDF]
* ''Abstract:'' In general, the less probable an event, the more attention we pay to it. Likewise, considering visual perception, it is interesting to regard important image features as those that most depart from randomness. This statistical approach has recently led to the development of adaptive and parameterless algorithms for image analysis. However, they require computer-intensive statistical measurements. Digital retinas, with their massively parallel and collective computing capababilities, seem adapted to such computational tasks. These principles and opportunities are investigated here through a case study: extracting meaningful segments from an image.
Changed line 12 from:
-> Paul Nadrag and Antoine Manzanera and Nicolas Burrus
to:
-> Paul Nadrag and [[http://www.ensta.fr/~manzaner|Antoine Manzanera]] and Nicolas Burrus
(:linebreaks:)
Changed lines 12-16 from:
->Paul Nadrag and Antoine Manzanera and Nicolas Burrus
->''In the Proceedings of Distributed Smart Cameras Workshop (DSC'06), 2006''
* [http://www.iti.tugraz.at/dsc06/CR/dsc06_p03_cr.pdf PDF]
* ''Abstract:'' The purpose of this work is to provide a robust vision-based input device. In our system, a programmable retina is looking at the user who sends commands by moving his hand. The fusion between the acquisition and the processing functions of the retina allows a close adaptation to the lighting conditions and to the dynamic range of the scene. Thanks to its optical input and massive parallelism, the retina computes ef&#64257;ciently the contours of the moving objects. This feature has nice properties in terms of motion detection capabilities and allows a dramatic reduction in the volume of data to be output of the retina. An external low-power processor then performs global computations on the output data, such as extreme points or geometric moments, which are temporally &#64257;ltered to generate a command.
to:
-> Paul Nadrag and Antoine Manzanera and Nicolas Burrus
''In the Proceedings of Distributed Smart Cameras Workshop (DSC'06), 2006''
* [http://www.iti.tugraz.at/dsc06/CR/dsc06_p03_cr.pdf PDF]
* ''Abstract:'' The purpose of this work is to provide a robust vision-based input device. In our system, a programmable retina is   looking at the user who sends commands by moving his hand. The fusion between the acquisition and the processing functions of the retina allows a close adaptation to the lighting conditions and to the dynamic range of the scene. Thanks to its optical input and massive parallelism, the retina computes ef&#64257;ciently the contours of the moving objects. This feature has nice properties in terms of motion detection capabilities and allows a dramatic reduction in the volume of data to be output of the retina. An external low-power processor then performs global computations on the output data, such as extreme points or geometric moments, which are temporally &#64257;ltered to generate a command.
Changed line 11 from:
''In the Proceedings of Distributed Smart Cameras Workshop (DSC'06), 2006''
to:
->''In the Proceedings of Distributed Smart Cameras Workshop (DSC'06), 2006''
Changed line 11 from:
In the Proceedings of Distributed Smart Cameras Workshop (DSC'06), 2006
to:
''In the Proceedings of Distributed Smart Cameras Workshop (DSC'06), 2006''
Changed line 11 from:
''In the Proceedings of Distributed Smart Cameras Workshop (DSC'06), 2006''
to:
In the Proceedings of Distributed Smart Cameras Workshop (DSC'06), 2006
Changed lines 8-12 from:
:Paul Nadrag and Antoine Manzanera and Nicolas Burrus
: ''In the Proceedings of Distributed Smart Cameras Workshop (DSC'06), 2006''
:* [http://www.iti.tugraz.at/dsc06/CR/dsc06_p03_cr.pdf PDF]
:* ''Abstract:'' The purpose of this work is to provide a robust vision-based input device. In our system, a programmable retina is looking at the user who sends commands by moving his hand. The fusion between the acquisition and the processing functions of the retina allows a close adaptation to the lighting conditions and to the dynamic range of the scene. Thanks to its optical input and massive parallelism, the retina computes ef&#64257;ciently the contours of the moving objects. This feature has nice properties in terms of motion detection capabilities and allows a dramatic reduction in the volume of data to be output of the retina. An external low-power processor then performs global computations on the output data, such as extreme points or geometric moments, which are temporally &#64257;ltered to generate a command.
to:
->Paul Nadrag and Antoine Manzanera and Nicolas Burrus
''In the Proceedings of Distributed Smart Cameras Workshop (DSC'06), 2006''
* [http://www.iti.tugraz.at/dsc06/CR/dsc06_p03_cr.pdf PDF]
* ''Abstract:'' The purpose of this work is to provide a robust vision-based input device. In our system, a programmable retina is looking at the user who sends commands by moving his hand. The fusion between the acquisition and the processing functions of the retina allows a close adaptation to the lighting conditions and to the dynamic range of the scene. Thanks to its optical input and massive parallelism, the retina computes ef&#64257;ciently the contours of the moving objects. This feature has nice properties in terms of motion detection capabilities and allows a dramatic reduction in the volume of data to be output of the retina. An external low-power processor then performs global computations on the output data, such as extreme points or geometric moments, which are temporally &#64257;ltered to generate a command.
Changed lines 1-78 from:
(:title Research:)
to:
(:title Research:)

!! Publications

!!! Computer Vision

!!!! [2006] Smart retina as a contour-based visual interface
:Paul Nadrag and Antoine Manzanera and Nicolas Burrus
: ''In the Proceedings of Distributed Smart Cameras Workshop (DSC'06), 2006''
:* [http://www.iti.tugraz.at/dsc06/CR/dsc06_p03_cr.pdf PDF]
:* ''Abstract:'' The purpose of this work is to provide a robust vision-based input device. In our system, a programmable retina is looking at the user who sends commands by moving his hand. The fusion between the acquisition and the processing functions of the retina allows a close adaptation to the lighting conditions and to the dynamic range of the scene. Thanks to its optical input and massive parallelism, the retina computes ef&#64257;ciently the contours of the moving objects. This feature has nice properties in terms of motion detection capabilities and allows a dramatic reduction in the volume of data to be output of the retina. An external low-power processor then performs global computations on the output data, such as extreme points or geometric moments, which are temporally &#64257;ltered to generate a command.

==== '''[2006] Adaptive Vision Leveraging Digital Retinas: Extracting Meaningful Segments''' ====
:Nicolas Burrus and Thierry Bernard
: ''In the Proceedings of Advanced Concepts for Intelligent Vision Systems International Conference (ACIVS'06), 2006''
:* [{{SERVER}}/files/research/burrus.06.acivs.meaningful_segments.pdf PDF]
:* ''Abstract:'' In general, the less probable an event, the more attention we pay to it. Likewise, considering visual perception, it is interesting to regard important image features as those that most depart from randomness. This statistical approach has recently led to the development of adaptive and parameterless algorithms for image analysis. However, they require computer-intensive statistical measurements. Digital retinas, with their massively parallel and collective computing capababilities, seem adapted to such computational tasks. These principles and opportunities are investigated here through a case study: extracting meaningful segments from an image.

==== '''[2005] Détection de segments significatifs sur rétine artificielle programmable (fr)''' ====
:Nicolas Burrus
: ''Rapport de stage de master''
:* [{{SERVER}}/files/research/burrus.05.report.segments_significatifs.pdf PDF]
:* ''Résumé:'' Ce travail se place à l’intersection de deux approches à priori indépendantes du traitement des images : une approche matérielle et algorithmique basée sur l’architecture cellulaire massivement parallèle des rétines ar ti&#64257;cielles ; et une approche a contrario statistiquement fondée de la perception cherchant à minimiser le nombre de paramètres nécessaires pour analyser les images. Nous nous proposons d’étudier la combinaison de ces deux univers à travers un opérateur simple et générique : l’extraction de segments signi&#64257;catifs.

==== '''[2004] Visualization and White Matter Fiber Tracking with Diffusion Tensor Magnetic Resonance Images''' ====
:Nicolas Burrus
: ''Internship report, Siemens Corporate Research (Princeton)''
:* ''Abstract:'' We propose to model DT-MRI fiber tracking using particle filters. The resulting algorithm can handle both noise and ambiguities raised by partial volume effects and crossing fibers. Combined with a simple clustering algorithm, distinct fiber branches can be explored without loss of tracking power.  Interesting results were obtained with a reasonable computation time.  Last, the flexibility of the framework makes it possible to add many parameters to the model, opening good perspectives for future improvements.
:* Patent application #20060229856 class: 703011000 (USPTO)

=== C++ related ===

==== '''[2004] Introduction to C++ metaprogramming''' ====
: Nicolas Burrus
: ''Tutorial''
:* [{{SERVER}}/files/research/burrus.04.report.metaprogramming.pdf PDF]
:* ''Abstract:'' This report aims at simplifying the discovery of the static C++ world. Mostly relying on Todd Veldhuisen Techniques for scientific C++ and Andrei Alexandrescu Modern C++ Design, we try to give to the novice metaprogrammer most of the basics notions he should learn, in a didactic way. The goal is also to make the reader work and think by himself before discovering already existing solutions, in order to facilitate the understanding.

==== '''[2003] A Static C++ Object-Oriented Programming (SCOOP) Paradigm Mixing Benefits of Traditional OOP and Generic Programming''' ====
:Nicolas Burrus, Alexandre Duret-Lutz, Thierry Geraud, David Lesage and Raphaël Poss
: ''In the Proceedings of the Workshop on Multiple Paradigm with OO Languages (MPOOL'03) Anaheim, CA, USA Oct. 2003''
:* [{{SERVER}}/files/research/burrus.03.mpool.scoop.pdf PDF]
:* ''Abstract:'' Object-oriented and generic programming are both supported in C++. OOP provides high expressiveness whereas GP leads to more efficient programs by avoiding dynamic typing. This paper presents SCOOP, a new paradigm which enables both classical OO design and high performance in C++ by mixing OOP and GP. We show how classical and advanced OO features such as virtual methods, multiple inheritance, argument covariance, virtual types and multimethods can be implemented in a fully statically typed model, hence without run-time overhead.
<blockquote>
<pre>
@article{burrus2003sco,
title={{A Static C++ Object-Oriented Programming (SCOOP) Paradigm Mixing Benefits of
author={Burrus, N. and Duret-Lutz, A. and Geraud, T. and Lesage, D. and Poss, R.},
journal={Workshop on multiple paradigm with OO languages. MPOOL},
volume={3},
year={2003}
}
</pre>
</blockquote>

==== '''[2002] Safe and efficient data types in C++ (Intègre)''' ====
:Nicolas Burrus
: ''LRDE technical report''
:* [{{SERVER}}/files/research/burrus.02.report.datatypes.pdf PDF]
:* ''Abstract:'' Using C++ builtin types is very unsafe as they are inherited from C types, which do not have overflow checking and have dangerous side effects and unexpected behaviors. Using intensive meta programming, it becomes possible to design safe data types with a minimal runtime overhead. As we want to be able to use existing algorithms, these types have to interact transparently with C++ builtin types. Primarily designed for Olena, a generic image processing library, our work needs to provide mechanisms to allow easy integration in generic algorithms.

=== Other ===

==== '''[2003] Theory of Evidence''' ====
:Nicolas Burrus and David Lesage
: ''LRDE technical report''
:* [{{SERVER}}/files/research/burrus_lesage.03.report.evidence.pdf PDF]
:* ''Abstract:'' The theory of evidence, also called Dempster-Shafer theory or belief functions theory, has been introduced by Glenn Shafer in 1976 as a new approach for representing uncertainty. Nowadays, this formalism is considered as one of the most interesting alternatives to Bayesian networks and fuzzy sets. This report makes an overview of both theoretical and implementation aspects of this theory. After a short survey of the historical motivations for this theory, we present its interesting properties through the Transferable Belief Model formalism. From a more practical point of view, we propose a review of the existing optimizations for facing the #P complexity of Dempster-Shafer computations. This report introduces a new, promising concept to compute repeated fusions: the delayed mass valuation. Finally, we present Evidenz, our general-purpose C++ library for designing efficient Dempster-Shafer engines.

==== '''[2001] Neural Networks: Multi-Layer Perceptron and Hop&#64257;eld Network''' ====
:Sylvain Berlemont, Nicolas Burrus, David Lesage, Francis Maes, Jean-Baptiste Mouret, Benoît Perrot, Maxime Rey, Nicolas Tisserand, Astrid Wang
: ''LRDE technical report''
:* [{{SERVER}}/files/research/csi2004.01.report.nn.pdf PDF]
:* ''Abstract:'' The ability of neural networks to derive meaning from complicated or imprecise data make them a powerful tool to extract hidden correlations between patterns or to recognize noised patterns. This report is dedicated to the study of Multi-Layer Perceptrons (MLP) and Hop&#64257;eld networks. In particular, two applications are detailed. MLP possibilities are illustrated through an image compression software and Hop&#64257;eld networks are studied through a character recognizer. For both applications, theoretical principles, heuristic and algorithmic improvements are discussed thanks to various experiments.
Changed lines 1-5 from:
Test complet

Comment c'est stocké ?

Faut voir !
to:
(:title Research:)
Changed lines 1-5 from:
Test
to:
Test complet

Comment c'est stocké ?

Faut voir !