CSC5651 Deep Learning In Signal Processing: Graduate Literature Project Milestone 1: Paper Selection

This page describes the first of five project milestones.

Instructions

Note: This assignment is only applicable to graduate students. Undergraduates should not complete this assignment and will be excused from the assignment within Canvas.

Graduate students in this course will work in teams of 2–3 to present two papers. You may form your own groups of 2–3, or if you prefer, the instructor can assign groups. This is a “launch” milestone for the project, which will be followed by a set of four milestones that is repeated for each of the two papers.

The papers selected should ideally have a strong relevance to the course topics, a clear and concise methodology, and significant impact on the field of study. They can be either foundational or recent works. Please provide a brief rationale for your paper selection when submitting your choices.

At least one of your papers should focus on a deep neural network architecture. In the early stages of a project, different classic and newer architectures are often evaluated to determine which one is the most promising for meeting the project’s overall goals.

Your second paper may also focus on an architecture, but you are also welcome to consider other topics related to the course such as data collection, data augmentation, error metrics, specific capabilities of MATLAB, PyTorch, TensorFlow, etc.

As you search for your papers, consider sources such as IEEE Xplore, the ACM Digital Library, and Google Scholar. Generally, your papers should be peer-reviewed, although exceptions are allowed especially for cutting-edge topics where a reputable but not peer-reviewed source is available. A good place to start is the library’s list of EECS library resources. If you have trouble accessing any of the suggested databases, please let the your instructor know as soon as possible they can help you find a solution.

The following list presents some recommended papers for the project. You are welcome to select from the list or choose your own.

CSC 5651 paper suggestions

Deliverable

For this “launch” milestone, you will need to post the following to this discussion board:

• Who you’d like to work with
• At least two papers that you’d like to present, with brief rationale for each

Requirements

• Review potential papers for this project, following the selection guidelines listed in the prior section.
• Submit your discussion board post by the posted due date. Be sure to address both aspects of the listed deliverable (i.e., team and papers).
  • If you have already formed a group, one member of your group may submit the post on the group’s behalf.
  • If you have not already formed a group, you may use this discussion board to find potential group members and discuss possibilities.
• Once your group has been formed, be sure to navigate to the People tab in the left-hand navigation menu to join the group within Canvas. Review the following guide as needed and reach out to your instructor with any questions.
  • How do I join a group as a student?
• Full credit will be awarded for timely submissions that fully address the assignment instructions.