Course Information

About

Over the past decade, deep learning models have achieved notable success across various domains such as computer vision, natural language understanding and speech processing, approaching or even surpassing human-level performances in many benchmark datasets. Yet deep learning keeps evolving and expanding into new frontiers. In this advanced seminar course, we’ll read and discuss a broad collection of papers on a wide variety of topics, including compositionality and systematic generalization, multimodal representation learning, memory and attention, graph neural networks, object-centric representation learning, dynamic networks, vision transformers, neural rendering, neural implicit representations, neuro-symbolic approaches, and deep implicit layers. The course also includes a project component, in which students will work alone or in pairs on a research topic covered in the class throughout the semester.

The structure of the course follows the format used in Colin Raffel’s COMP790 course at University of North Carolina and Alec Jacobson's CSC2521 course at the University of Toronto. Each lecture students will play a different role that defines how they will read the paper and focus on distinct points of view while preparing for the in-class discussion. See Presentations page for the details. This process is chosen to provide multiple perspectives, a thorough understanding of the concepts, and more importantly way more fun. The aim of this seminar is to bring students to the state of the art in this exciting field.

This class is taught by Aykut Erdem, and intended for graduate students and ambitious undergraduates with a research experience. To get the most out of the class, students should have a strong knowledge in deep learning (such as COMP541 or COMP547) and good programming skills. If you have doubts whether you meet these requirements, please consult the instructor in advance.

Instruction style: During the semester, students are responsible for studying and keeping up with the course material outside of class time. These may involve reading particular book chapters, papers or blogs and watching some video lectures.

Time and Location

Lectures: Monday and Wednesday at 11:30-12:40 (CASE Z25)

Communication

The course webpage will be updated regularly throughout the semester with lecture notes, presentations, and important deadlines. All other course related communications will be carried out through Blackboard.

Course Requirements and Grading

Grading for COMP550 will be based on

  • Course Project (presentations and reports) (48%),
  • Paper Presentations (52%),

Reference Books

  • Ian Goodfellow, Yoshua Bengio, and Aaron Courville, Deep Learning, MIT Press, 2016 (draft available online)

Schedule

Week Topic Assignments
Sep 27-29 Introduction to the course
Oct 4-6 Compositionality and systematic generalization
Oct 11-13 Multimodal representation learning
Oct 18-20 Graph neural networks Project proposals due
Oct 25-27 Object-centric representation learning
Nov 1-3 Neuro-symbolic approaches
Nov 8-10 Neural implicit representations
Nov 15-17 Winter break
Nov 22-24 Project progress presentations Project progress reports due
Nov 29-Dec 1 Dynamic networks
Dec 6-8 Neural rendering
Dec 13-15 Memory and attention
Dec 20-22 Vision Transformers
Dec 27-29 Deep implicit layers
Jan 3-5 Final Project Presentations Final project reports due
Detailed Syllabus and Lectures