NEW CLASS IE 590, FALL 2019:
DEEP LEARNING AND MACHINE VISION*
Dr. Juan P. Wachs (email@example.com)
This course combines practically relevant methods, tools and applications related to deep learning in computer vision. Specifically, we will look at how data science principles are applied to computer vision through: (1) teaching neural networks including deep configurations with recent advances such as Generative Adversarial Networks (GANs), Long short-term memory (LSTMs), Style Transfer, and Autoencoders.
Program Overview (in a nutshell):
1. Introduction to computer vision and deep learning.
2. Linear Classification, Loss Functions, Neural Networks and Backpropagation
3. CNNs and Recurrent Neural Networks
4. Feature Representation and Adversarial Networks.
5. Deep Reinforcement Learning.
6. Style Transfer, GANs, RNNs, LSTM and Network Visualization.
Individual Assignments: Three homeworks will be assigned during the semester.
Project Assignments: Students will form a small team and build an application (or model) of a computer vision deep learning based system.
Midterm: An in class midterm will be given to the students in class.
Course Requirements: Calculus, linear algebra, probability, Python programming experience (C/C++ is also a plus).