新IT卓越大讲堂 No.16 期
Course on Deep learning for computer vision
This course covers basic and advanced concepts of deep learning applied to computer vision. We will first introducesupervised learning methods for deep neural architectures, in particular, convolutional neural networks and recurrent neural networks. We willthen present recent techniques related to generative modeling, weakly supervised learning and reinforcement learning. Specifically, the following topics will be covered:
· Deep learning: motivation and historical context, multilayer networks, convolutional networks;
· Training: backpropagation, gradient descent, regularization, data augmentation;
· Recurrent networks: gradient propagation, LSTM networks, multi-scale networks, applications;
· Generative models: autoencoders, generative adversarial networks, applications;
· Learning with reduced supervision: weakly and semi supervised learning, attention models, curricular learning;
· Reinforcement learning: Markov decision process, dynamic programing, temporal difference learning, Monte Carlo methods.
Christian Desrosiers obtained a Ph.D. in Computer Engineering from Polytechnique Montreal and was a postdoctoral researcher at the University of Minnesota. In 2009, he joined ETS, University of Quebec, as professor in the Departement of Software and IT Engineering. Prof. Desrosiers is codirector of the Laboratoired’imagerie, de vision et d’intelligenceartificielle (LIVIA) and is a member of the REPARTI research network. His main research interests focus on machine learning, image processing, computer vision and medical imaging. He has published over 100 peer-reviewed papers in these fields and has been on the scientific committee of several important conferences like European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (EML-PKDD).