Program

 PROGRAM

Monday April 19

 

 

 

12 pm Paris time
6 pm Beijing time
6 am New-York time

 Welcome and introduction
Bruno Montcel and Thomas Grenier, CREATIS, Lyon
   
12.50 pm Paris time
6.50 pm Beijing time
6.50 am New-York time		 Introduction to machine learning
Odyssée Merveille and Emmanuel Roux, CREATIS, Lyon
Basics, train and test sets, metrics, over and under fitting, ... - video and slides
2.30 pm Paris time
8.30 pm Beijing time
8.30 am New-York time

Questions/ discussion 20'
Chair: Thomas Grenier
   
3 pm pm Paris time
9 pm Beijing time
9 am New-York time		 Basics in deep learning 1
Pierre-Marc Jodoin, Sherbrooke University
Perceptron and multi-layer perceptron, stochastic gradient descent, learning rate, logistic regression, activation function, regularization (L1/L2/dropout/early stopping), … video and slides
4.05 pm Paris time
10.05 pm Beijing time
10.05 am New-York time		 Questions/ discussion 20'
Chair: Odyssée Merveille
   
4.25 pm Paris time
10.25 pm Beijing time
10.25 am New-York time		 Musical blind test
by Yohann
   
5.10 pm Paris time
11.10 pm Beijing time
11.10 am New-York time		 Basics in deep learning 2
Christian Desrosiers, ETS Montréal
Weights initialization, forward and backward propagation, batch size, convolution neural nets (CNN), feature maps, pooling, pretraining and transfer learning, applications - video and slides
 6.35 pm Paris time
12.35 am Beijing time
12.35 pm New-York time		 Questions/ discussion 20'
Chair: Emmanuel Roux
   
Tuesday April 20

 

 
1pm Paris time
7 pm Beijing time
7 am New-York time		 Advanced concepts in deep learning 1
Olivier Bernard, CREATIS, Lyon
Common CNN architectures for classification (VGGNet, ResNEt, ...) and localization (FasterRCNN, Yolo) and segmentation (encoder-Decoder, U-Net, ENet, ...) - video and slides
2.15 pm Paris time
8.15 pm Beijing time
8.15 am New-York time		 Questions/ discussion 20'
Chair: Nicolas Duchateau
   
2.35 pm  Paris time
8.35 pm Beijing time
8.35 am New-York time		 Escape game presentation and instructions
by Emmanuel Roux
   
3.05 pm Paris time
9.05 pm Beijing time
9.05 am New-York time		 Escape game n°1
  break
   
4.20 pm Paris time
10.20 pm Beijing time
10.20 am New-York time

Introduction by M Sdika - video and slides

Advanced concepts in deep learning 2
Narine Kokhlikyan, Facebook, KIT, Karlsruhe
Visualization, explainability, interpretability - video and slides
5.50 pm Paris time
11.50 pm Beijing time
11.50 am New-York time

 

Questions/ discussion 20'
Chair: Michaël Sdika
   
  Meal break
   
7 pm Paris time
1 pm New-York time		 Hands-on session 1.1: classification from machine to deep learning
Nicolas Duchateau, Thomas Grenier, Olivier Bernard
   
Wednesday April 21

 

 
9 am Paris time
3 pm Beijing time		 Hands-on session 1.2: classification from machine to deep learning
Nicolas Duchateau, Thomas Grenier, Olivier Bernard
   
1 pm Paris time
7 pm Beijing time
7 am New-York time		 Generative and adversarial methods for medical imaging
Anirban Mukhopadhyay, Technische Universität Darmstadt
GANs, autoencoders and their training - video

2.15 pm Paris time
8.15 pm Beijing time
8.15 am New-York time

 Questions/ discussion 20'
Chair: Carole Lartizien
   
2.35 pm Paris time
8.35 pm Beijing time
8.35 am New-York time		 Yoga and good posture (TBC)
   
3.20 pm Paris time
9.20 pm Beijing time
9.20 am New-York time		 Bayesian neural network - Uncertainty
Rémi Emonet, Laboratoire Hubert Curien, Saint-Etienne
Types of uncertainty, classifier calibration, probabilistic modeling, Bayesian Neural Networks, dropout and variations, ensemble methods - video and slides
 4.35 pm Paris time
10.35 pm Beijing time
10.35 am New-York time

 

Questions/ discussion 20'
Chair: Fabien Millioz
  break
   
5.10 pm Paris time
11.10 pm Beijing time
11.10 am New-York time		 Round Table - Topics : Sharing medical data for machine learning:  local/global initiatives ; issues ; Regulation
  Meal break
   
7 pm Paris time
1 pm New-York time		 Hands-on session 2.1: Variational Autoencoder
Pierre Marc Jodoin, Nathan Painchaud
Auto-encoders, convolutional auto-encoders, variational auto-encoders, latent spaces
   
Thursday April 22

 

 
9 am Paris time
3 pm Beijing time		 Hands-on session 2.2: Variational Autoencoder
Pierre Marc Jodoin, Nathan Painchaud
Auto-encoders, convolutional auto-encoders, variational auto-encoders, latent spaces
   
1pm Paris time
7pm Beijing time
7am New-York time		 Deep learning for inverse problems
Nicolas Ducros, CREATIS laboratory
Image reconstruction, computational optics, convolutional neural networks, interpretable networks - video and slides
2.15pm Paris time
8.15pm Beijing time
8.15am New-York time		 Questions/ discussion 20'
Chair: Suzanne Bussod
  break
   
2.50 pm Paris time
8.50 pm Beijing time
8.50 am New-York time		 Deep learning for image acquisition and reconstruction
Ruud Van Sloun, Eindhoven University of technology
MRI/US deep learning based acquisition and reconstruction, unfolding method, hybrid data driven approach, end-to-end optimization - Video and slides
3.50 pm Paris time
9.50 pm Beijing time
9.50 am New-York time		 Questions/ discussion 20'
Chair: Odyssée Merveille
  break
   
4.30 pm Paris time
10.30 pm Beijing time
10.30 am New-York time		 Poster session
  Meal break
   
7 pm Paris time
1 pm New-York time		 Hands-on session 3.1: reconstruction using deep learning
Nicolas Ducros
   
Friday April 23

 

 
9 am Paris time
3 pm Beijing time		 Hands-on session 3.1: reconstruction using deep learning
Nicolas Ducros
   
1 pm Paris time
7 pm Beijing time
7 am New-York time		 Geometric deep learning
Hervé Lombaert, ETS, Montréal
Spectral coordinates and representation, spectral deep learning, brain surface matching and parcellation - video and slides
2.15 pm Paris time
8.15pm Beijing time
8.15am New-York time		 Questions/ discussion 20'
Chair: Pierre-Marc Jodoin
  break
   
2.50 pm Paris time
8.50 pm Beijing time
8.50 am New-York time		 Escape game n°2
  break
   
4.05 pm Paris time
10.05 pm Beijing time
10.05 am New-York time

Weakly supervised deep learning
Ismail Ben Ayed, ETS, Montréal and Jose Dolz, ETS, Montréal
Weakly supervised segmentation, constrained CNN losses, semantic segmentation, semi-supervised learning - Video I Ben Ayed, video J Dolz and slides
5.30 pm Paris time
11.30 pm Beijing time
11.30 am New-York time

 

Questions/ discussion 20'
Chair: Christian Desrosiers
   
5.50 pm Paris time
11.50 pm Beijing time
11.50 am New-York time		 School conclusions
  Meal break
   
7 pm Paris time
1 pm New-York time

Hands-on session 4.1: Segmentation using deep learning
Thomas Grenier, Olivier Bernard, Emmanuel Roux
   
Saturday April 24

 

 
9 am Paris time
3 pm Beijing time		 Hands-on session 4.2: Segmentation using deep learning
Thomas Grenier, Olivier Bernard, Emmanuel Roux

 
Online user: 3 Privacy
Loading...