Date |
Topic |
Recommended readings |
Slides |
2016-06-02, 10am, room A103 |
Introduction to Neural Networks |
Bishop, Chapter 5 |
slides (Pablo Mesejo, Stéphane Lathuilière) |
2016-06-10, 10am, room F107 |
Introduction to Neural Networks |
Bishop, Chapter 5 |
slides (Stéphane Lathuilière, Pablo Mesejo) |
2016-07-01, 10am, room F107 |
Introduction to Convolutional Neural Networks |
Goodfellow et al., Deep Learning, Chapter 9
Visualizing and Understanding Convolutional Networks |
slides1 slides2 (Vicky Kalogeiton, Shreyas Saxena) |
2016-09-22, 10am, room A103 |
Unsupervised Deep Learning |
Generative Adversarial Networks, Auto-Encoding Variational Bayes, Tutorial on Variational Autoencoders |
slides1 slides2 slides3 (Daan Wynen, Jakob Verbeek) |
2016-09-30, 10am, room F108 |
Introduction to Recurrent Neural Networks |
Graves, Supervised Sequence Labelling with Recurrent Neural Networks, Ch 3-3.2.4 and Ch 4
Goodfellow et al., Deep Learning, Chapter 10-10.3 |
slides (Marco Pedersoli, Thomas Lucas) |
2016-11-4, 10am, room F107 |
Deep Feedforward Networks |
Goodfellow et al., Deep Learning, Chapter 6 |
slides (Benoit Massé, Dionyssos Kounades-Bastian) |
2016-11-17, 15’30, room F107 |
Regularization for Deep Learning |
Goodfellow et al., Deep Learning, Chapter 7 |
slides (Gildas Mazo) |
2016-12-13, 10am, room F108 |
Optimization for Training Deep Models |
Goodfellow et al., Deep Learning, Chapter 8 |
slides (Henrique Morimitsu) |
2017-01-25, 10am, room F107 |
Deep Learning Libraries |
– |
slides (Konstantin Shmelkov, Pauline Luc, Thomas Lucas, Vicky Kalogeiton, Stéphane Lathuilière) |
2017-10-03, 3pm, room C208 |
Deep Learning Papers |
Generative Adversarial Nets, Realtime Multi-Person 2D Pose Estimation using Part Affinity Fields, Deeply-Learned Part-Aligned Representations for Person Re-Identification, Pose-driven Deep Convolutional Model for Person Re-identification, Disguised Face Identification (DFI) with Facial KeyPoints using Spatial Fusion Convolutional Network, Deep Forest: Towards An Alternative to Deep Neural Networks |
Sylvain Guy, Guillaume Delorme, Radu Horaud, Vincent Drouard, Stéphane Lathuilière, Benoit Massé |
2017-10-17, 3pm, room C208 |
Deep Learning Papers |
Detect to Track and Track to Detect, Interpretable Convolutional Neural Networks, Beyond triplet loss: a deep quadruplet network for person re-identification, Unsupervised Image-to-Image Translation Network, A Brief Survey of Deep Reinforcement Learning |
Sylvain Guy, Xavier Alameda, Benoit Massé, Stéphane Lathuilière, Pablo Mesejo |
2017-11-07, 2pm, room F108 |
Deep Learning Papers |
Deformable Convolutional Networks, Encouraging LSTMs to Anticipate Actions Very Early, Genetic CNN, Rethinking Reprojection: Closing the Loop for Pose-Aware Shape Reconstruction From a Single Image, Learning a Policy from Streaming Videos with Reinforcement Learning, Learning From Noisy Labels With Distillation |
Radu Horaud, Guillaume Delorme, Benoit Massé, Vincent Drouard, Yutong Ban, Stéphane Lathuilière |
2017-11-21, 3pm, room F108 |
Deep Learning Papers |
Following Gaze in Video, Sliced Wasserstein Distance for Learning Gaussian Mixture Models, Data Augmentation Generative Adversarial Networks, Bayesian GAN, SSH: Single Stage Headless Face Detector, SafetyNet: Detecting and Rejecting Adversarial Examples Robustly |
Benoit Massé, Yutong Ban, Stéphane Lathuilière, Hongliang Lu, Sylvain Guy, Guillaume Delorme |
2017-12-05, 3pm, room F108 |
Deep Learning Papers |
Are GANs Created Equal? A Large-Scale Study, Deep clustering: Discriminative embeddings for segmentation and separation, How far are we from solving the 2D & 3D Face Alignment problem?, StarGAN: Unified Generative Adversarial Networks for Multi-Domain Image-to-Image Translation, Are You Talking to Me? Reasoned Visual Dialog Generation through Adversarial Learning |
Stéphane Lathuilière, Laurent Girin, Benoit Massé, Guillaume Delorme, Yutong Ban |
2018-01-23, 3pm, room C208 |
Deep Learning Papers |
Mask R-CNN, Domain-Adversarial Training of Neural Networks, Camera Style Adaptation for Person Re-identification, Gradients explode – Deep Networks are shallow – ResNet explained |
Yutong Ban, Guillaume Delorme, Guillaume Delorme, Óscar Gómez |