Course Description 

This graduate course introduces basic and advanced topics in deep learning and its application in computer vision. We will discuss deep learning architectures including CNN, RNN, LSTM, and transformers, optimization and regularization of neural networks, and discriminative and generative models for solving different computer vision tasks, including visual perception for image, video, and 3D data, visual content generation, and recent advances in the field.  

The course includes 3 hours of lectures per week. Assignments include both written exercises and programming exercises. A course project with a final presentation and a report is required for each student.      

Prerequisites 

We require students to have prior knowledge in linear algebra, statistics, and programming.