CSC2231H S LEC0101 20231: Visual and Mobile Computing

Course Description

Visual computing tasks such as 2D/3D graphics, image/video processing, sensor data processing, and image understanding are important aspects of modern computer systems ranging from sensor-rich smartphones to robotics to large data centers. These workloads demand exceptional system efficiency and this course examines the key ideas, techniques, and challenges associated with the design of parallel (and heterogeneous) systems that serve to accelerate visual computing applications. This course will cover how these techniques are applied in modern robotics, autonomous vehicles, computational photography, and smart sensor nodes. This course will also explore how to synergistically co-design and co-optimize across layers of a computing stack, including applications, programming frameworks, compilers, OS, and hardware architecture. This course is intended for graduate and advanced undergraduate-level students interested in architecting efficient future platforms for image processing, computer vision, and mobile robotics and students seeking to understand these application domains and develop scalable algorithms for these platforms. This course will comprise lectures, paper readings, and a course-length project. Background in robotics, visual computing, or machine learning is not required for this course.

Course Project

You need to make a semester-long project. You can form teams of any size between 1-3, and pick any project of your choice. As part of the course project, you are expected to

• Submit a project proposal (due the 9th of February),
• Write a final report (due at the end of the quarter),
• Present your results with a poster at the end of the quarter (tentative, as we need to figure out the logistics on this right now)

Project Proposal

The project proposal should contain the following things in a 1-2 page proposal. 

• Goal: what are you trying to do
• Background: some background on the topics in question
• Milestones: 
  • Mid-semester milestone
  • End-of-semester milestone
• Methodology: how do you plan to do this research. 
• Prior work: A brief discussion of the latest research that looks at similar problems/techniques. 

We recommend starting working on the project as soon as you can. You are free to adapt your research as the course project if it aligns with any topics in the course (computer vision, robotics, performance optimization, edge-computing, machine learning, architecture, system optimization).

Presentation

We will hold a poster session at Behan Atrium. We will release details later.

Peer Review

This yeat, our course has 47 enrolled students with 26 courses projects. We want students also have participations in the paper review process. Each project will be reviewed by 5-8 students. We use OpenReview to manage the project submission and review.