Didier COQUIN

University Professor, Annecy University Institute of Technology, Networks and Telecommunications Department

Contact

Email: didier.coquin –@– univ-smb.fr

Phone: +33(0) 4 50 09 65 46 – Fax: +33(0) 4 50 09 65 ..

Office: A118

Address: LISTIC – Polytech Annecy-Chambéry, PO Box 80439, 74944 Annecy le Vieux Cedex, France

Research topics.

Since July 2018, my work has been part of the AFuTé (Apprentissage, Fusionet Télédétection)group. My research focuses on the general theme of using fusion methods for decision support in a context of imperfect, uncertain, or even complementary data. It is structured around three areas:

  • Interval approach and thick fuzzy sets
  • Credibility approaches
  • Merging of additional information

From September 2014 to July 2018: Facilitator of the Combination and Decision (CoDe) group.

  • My work focuses on combining subclassifiers to recognize tree species based on a photo of one of their leaves. The recognition task is complicated by the fact that there is significant intra-class variability and inter-class similarity. Belief function theory is applied to the output of each subclassifier (random forests) in order to refine the results by reducing the effect of uncertainty on leaf characteristics. New work has been developed on the architecture of the fusion system so that it can adapt to the different modalities (leaf, bark, fruit, flower, etc.) present at the input. We have developed a hierarchical fusion system by combining the different outputs of the system using belief function theory.
  • I am also working on merging multi-sensor information using belief function theory. In one of the applications developed, we combined the outputs of a Takagi Sugeno fuzzy system with constant conclusions to a credential combination system that removes ambiguities about the color detected by the cameras. In a second application, we improved the perception of the NAO humanoid robot by adding 2D and 3D cameras in its immediate environment. Each camera provides information in the form of an image. After processing and extracting the useful information, we merged these different sources to make a decision. Combining them using the Dempster-Shafer operator increases the rate of correct decisions.
  • In collaboration with the EDYTEM laboratory, we have been working on multispectral image processing and data spatialization for the characterization of organic matter in sediment cores. We have VNIR and SWIR images with different resolutions. The images must first be recalibrated, and since the wavelengths of these images do not overlap, the fusion problem remains complex. Four methods have been developed. The first consists of combining hyperspectral and conventional methods to create quantitative predictive models. The second allows for the spatial recalibration of the different hyperspectral images at the lowest resolution. The third focuses on merging these images at the highest resolution. Finally, the last method focuses on deposits present in sediments (laminae, floods, etc.) to add a temporal dimension to our studies.

January 2013 to August 2014: Facilitator of the Information Fusion for Decision Support (FIAD) group.

From Sept. 2010 to Sept. 2012: Coordinator of the EFIX theme: Extractionand Fusionof InformationfromComplexSystems

  • Main contribution to the "Fusion Method" research area, particularly in image processing applications. My work focuses on developing numerical and symbolic criteria for comparing and evaluating image processing techniques.
  • Contribution also on the performance evaluation of information fusion systems. System performance is based on the definition of a "mission-quality" criterion associated with each block (Extraction, Representation, Fusion, Decision) of the system.

Publications

listed in HAL

Teaching at the Annecy University Institute of Technology, in the Network and Telecommunications Bachelor's Degree program

  • Responsible for first-year computer science courses, tutorials, and practicals (Resource 107: Programming Fundamentals). The language used is Python.
  • Head of SAé Data Processing 1st year (Creating IT tools and applications for R&T) The language used is Python.
  • Course Coordinator and Lab Instructor for First-Year Computer Science (Resource 208: Analysis and Processing of Structured Data) in Python
  • Head of Telecommunications Courses, Tutorials, and Practical Work for First-Year Students (Resource 205: Signals and Systems for Transmission)
  • Head of Telecommunications Courses, Tutorials, and Practical Work for Second-Year Students (Resource 305: Digital Transmission Channels)
  • Course Coordinator, Probability and Statistics Tutorials, 2nd year (Resources 405)
  • Participation in tutorials/practical sessions on other resources taught by colleagues in the department.
  • Head of the DATA931 module: Machine Learning: 5th year of engineering school at Polytech Annecy.

Responsibilities within the R&T department

  • Head of the R&T department since July 2022.
  • Head of Further Education from July 2020 to July 2022.
  • Head of the R&T department from July 2014 to July 2020.

Research Activities.

Activities / research groups.

December 15, 2022: Organization with Su Ruan and Mauro Dalla Mura of the meeting: Multimodal Machine Learning and Information Fusion: 3rd Edition (GDR ISIS – Theme B)

January 19, 2022: Organization with Su Ruan and Mauro Dalla Mura of the meeting: Multimodal Machine Learning and Information Fusion: 2nd Edition (GDR ISIS – Theme B)

In October 2013, I organized a meeting with Thierry Denoeux on information fusion within Theme A of the ISIS GRD: Contribution of uncertainty theory to information fusion (GDR ISIS – Theme A).

July 2007 to November 2012: Member of the steering committee for Action 3: SCATI (Complex Systems for Image Analysis and Processing) with Rémy Mullot and Régis Clouard, from Theme B (Image and Vision) of the ISIS GDR.

Reviewer for the following journals: Fuzzy Sets and Systems, IEEE Transactions on Fuzzy Systems,IEEE Transactions on Pattern Analysis and Machine Intelligence, IEEE Transactions on Image Processing, IEEE Transactions on Instrumentation & Measurement, Pattern Recognition Letters, International Journal of Multimedia Information Retrieval, Computational Intelligence and Neuroscience, Computers in Biology and Medicine, Computers and Electronics in Agriculture, Biosystems Engineering, etc.

Reviewer for ACIVS conferences (2004, 2005), SPIE ISOT (2005, 2006, 2007, 2008), IEA/AIE (2006), IEEE ISOT (2009, 2010), HSI (2015), CIVEMSA (2017), CTIC (2017, 2019), FUZZ-IEEE (2019)

 

Research Projects

2018-2022: Participation inthe INTERREG France-Switzerland CIME program led by Frédéric POURRAZ.

2016-2020: LISTIC manager forthe ANR ReVeRIES (Reconnaissance de Végétaux Récréative, Interactive et Educative sur Smartphone - Recreational, Interactive, and Educational Plant Recognition on Smartphones) project led by LIUM (University of Le Mans). This project has two objectives: firstly, plant recognition using smartphone applications based on digital image analysis and data fusion; secondly, the development of a fun and educational application for training and creating "citizen science" inventories. http://reveries-project.fr

To download the JAVA applet (RF-EkNN): Evidential two-step tree species recognition approach from leaves and barks, click here: https://projects.listic.univ-smb.fr/reveries/RF_EkNN.zip

2010-2014: LISTIC manager forthe ANR ReVeS (Plant Recognition for Smartphone Interfaces) project led by LIRIS (Lyon). We are interested in merging heterogeneous information (data extracted from tree leaves, GPS data, phytogeographic data) to aid decision-making (providing the app user with the five most likely trees based on a photo of one of its leaves). http://liris.cnrs.fr/reves/

 

Thesis in progress:

…

Theses defended:

PAGNIER Fanny: "Decision support methods for characterizing the vigilance required in avalanche terrain," co-supervisors: Frédéric Pourraz, Hervé Verjus, Gilles Mauris, defended on June 19, 2023.

JACQ Kévin: "Multispectral image processing and data spatialization for the characterization of organic matter in natural solid phases," co-director: Bernard Fanget (EDYTEM). Defended on November 27, 2019.

BEN AMEUR Rihab: "Multimodal fusion for tree species recognition," co-director: Lionel Valet, defended on June 4, 2018.

NGUYEN Thanh Long: "Multi-sensor information fusion for controlling the NAO robot," co-supervisor: Reda Boukezzoula, defended on April 5, 2017.

CERUTTI Guillaume: "Segmentation and interpretation of natural images for the identification of tree leaves on smartphones," co-supervisors: Laure Tougne and Antoine Vacavant, defended on November 21, 2013.

LAMALLEM Abdellah: "Performance evaluation of an information fusion system: application to 3D image processing." Co-supervisor: Lionel Valet, defended on July 17, 2012.

IONESCU Bogdan: "Symbolic characterization of image sequences: application to animated films," thesis advisor Patrick Lambert, defended on May 7, 2007.

CHEHADEH Yousra: "Local operators for rectangular and parallelepiped meshes: application to image analysis," thesis advisor Philippe Bolon, defended on October 1, 1997.

Program Committee.

Member of the Program Committee for the annual SPIE ISOT (International Symposium on Optomechatronic Technologies) conference, Session: Computer Vision Systems, from 2005 to 2008, and co-chair of this conference. In 2009, this conference became an IEEE conference. In 2010, this conference will take place in Toronto, Canada.

Organization of the special session: Information Fusion Systems (SPS 30) at the IPMU 2010 conference in Dortmund, Germany, then IPMU 2012 in Catania, Italy, IPMU 2016in Eindhoven, IPMU 2018 in Cadiz, andIPMU 2020 in Lisbon (Portugal): "Belief function theory and Its Applications."

University authorities.

Elected member of CODUS (2012-2016).

LISTIC representative at ED-SISEO (2012-2016).

Elected member of the Commission of Specialists, 61st/63rd sections, then 27th/61st sections since 1995.

Elected member of the Board of Directors of the Annecy University Institute of Technology from 2002 to 2005.

Member of the Board of Directors of the Annecy University Institute of Technology from July 2014 to July 2020

Elected member of the Board of Directors of the Annecy University Institute of Technology since July 2021.

Member appointed to the Commission of Specialists, 61st section of the INPG from 1998 to 2005.

National Coordinator of the Telecom Group for the 28 R&T departments in France and Metropolitan France from 2003 to 2009.

Professional background.

  • Qualification to supervise research – December 2007 – University of Savoie.
  • Doctorate in Signal Processing and Telecommunications – 1991 – University of RENNES 1
  • Master's degree in Signal Processing and Telecommunications – 1987 – University of RENNES 1
  • Master's degree in EEA – 1986 – University of RENNES 1
  • Bachelor's degree in Economics and Business Administration – 1985 – University of Rennes 1
  • DEUG A: Mathematics and Physics – 1984 – University of RENNES 1.

Professional experience.

  • University professor, Annecy University Institute of Technology, R&T department, since September 1, 2012.
  • Lecturer, Annecy University Institute of Technology, R&T department, from September 1, 1998, to August 31, 2012
  • Lecturer, Annecy University Institute of Technology, GEII department, from October 1, 1992, to August 31, 1998
  • Research Engineer, FAROS Company, from July 1, 1992, to September 30, 1992
  • Assistant Professor – ENSSAT Lannion, University of Rennes 1 from October 1, 1990, to June 30, 1992
  • ALER – ENSSAT Lannion, University of RENNES 1 from October 1, 1988, to September 30, 1990
  • Associate Assistant – ENSSAT Lannion, University of RENNES 1 from September 1, 1987, to September 30, 1988