Trusted Coalitions and Efficient Distributed Deep Learning for Improved Image-based Decisions


The Montreal Chapter of the IEEE Signal Processing Society, in collaboration with STARaCom, cordially invites you to attend the following “pizza talk”, to be given by Prof. Benoit Macq (IEEE Fellow), from UCLouvain, Belgium.

  Date and Time




  • 3480 Univerisity
  • Montreal, Quebec
  • Canada H3A 0E9
  • Building: McConnell Eng.
  • Room Number: MC603
  • Click here for Map
  • Prof. Benoit Champagne

    ECE Department, McGill University

  • Starts 29 May 2019 01:55 PM
  • Ends 20 June 2019 01:55 PM
  • All times are Canada/Eastern
  • No Admission Charge
  • Register


Benoit Macq

Benoit Macq of UCLouvain, Louvain-la-Neuve, Belgium


Trusted Coalitions and Efficient Distributed Deep Learning for Improved Image-based Decisions

Image is the reference modality for computer-assisted decision in medicine, in urban surveillance and for many life interactive user experiences, like indoor amusement rides. These decision systems require shared and trusted models built from supervised learning among the community of users. In many cases, the learning can be accelerated by constituting coalitions of partners, each of them working on similar data. In this case, distributed learning allows the members of the coalition to train and share a model without sharing the data used to optimize this model.  A security architecture will be presented and analyzed from an Information Theory point-of-view which guarantees the preservation of data privacy to each member of the coalition, and a fair usage of the shared model by using adequate encryption, watermarking and blockchain mechanisms. We will demonstrate its effectiveness in the case of the distributed optimization of a Deep Learning Convolutional Neural Network trained on medical images. This architecture can be extended by not only sharing data in the coalition but also sharing analysis procedures. Concrete examples will be given in projects in the field of image-guided radiotherapy, urban surveillance and interactive dark rides.


Prof. Benoit Macq received is Diploma in Electrical Engr. from UCLouvain, Louvain-la-Neuve, Belgium in 1984. He did his military service at the Royal Military School on development of Lidar interferometry. He did his PhD thesis on digital TV compression in the frame of the RACE program of the European Union contributing to JPEG and MPEG groups. He received his PhD degree from UCLouvain- Belgium in 1989. He has been researcher in Philips Research Laboratory Belgium in 1990 and 1991 developing wavelet-based compression algorithms and contributing to the JPEG-2000 standard. Benoit Macq is Professor at Polytechnic School of UCLouvain since 1993. He has been visiting Professor at Ecole Polytechnique Fédérale de Lausanne, at the Massachusetts Institute of Technology, Boston and he is currently visiting professor at McGill University in Montreal. His main research interests are image compression, image watermarking, image analysis for medical and immersive communications. Benoit Macq was General Chair of IEEE ICIP2011 in Brussels. He was associate editor of IEEE Trans on Multimedia, IEEE Trans on Image Processing, Guest Editor for the Proceedings of the IEEE. He is the founder of the “Journal of Multimodal Interfaces” edited by Springer Verlag. He has been chair of several European projects, including a Network of Excellence on Multimodal Interactions (SIMILAR project) and an Integrated Proposal, EDcine which paved the way of the technology of digital cinema in European Union. He is the co-founder of 11 spin-off companies. Benoit Macq is Fellow Member of the IEEE and Member of the Royal Academy of Science of Belgium. He is currently Senior Editor Associate of IEEE Trans on Image Processing.


Refereshments will be served.