Skip to content

sirutBuasai/advanced_gen_modeling_mqp_2022_2023

 
 

Repository files navigation

"Advanced Applications of Deep Generative Models" MQP

Advisor: Elke Rundensteiner (rundenst@wpi.edu), Graduate Mentors: Walter Gerych (wgerych@wpi.edu), Joshua DeOliveira (jcdeoliveira@wpi.edu)

Incomplete, messy data is ubiquitous in real-world applications. The focus of this MQP is on generative methods to combat this issue and “fill in” incomplete data. Specifically, we will be focusing on the use of generative modeling to mitigate these issues in the Human Activity Recognition (HAR) domain.

This is your MQP and provides an opportunity to show off your skills. Whether your goal is industry or academia, the MQP is useful for applications. The Advisor and Mentor serve as guides during the MQP. While we will provide support and a general direction, we want all of you to bring your own ideas and seek out new directions.

Acknowledgments: This page borrows heavily from resource created by Prof. Tlachac.


Expectations

  • 15 to 20 hours per week of work
  • Two formal meetings per week: one with advisor and one with mentor. Prepare slides for both meetings
  • At least one peer meeting per week
  • Stay in communication with team (slack)
  • Teach team members as requested
  • Experience with being term leader and/or note taker. Leader ensures all team members are on time with term deliverables. Note taker records ideas for each meeting to share with team

Deliverables

  • MQP report, improved each term
    • includes literature review
    • write one part, edit one part
    • cite all sources
    • write succinctly (less can be more!)
    • include appendix with your term contributions
    • Potential to extend MQP report to full conference paper
  • Weekly update slides and presentation (for both advisor meeting and mentor meeting)
    • Include setup (don’t assume we remember!), what you did, and what you plan to do
    • Always include your conclusions and/or proposed next steps as appropriate
    • Complete slides and include written summary if unavailable to meet
    • Send draft of advisor slides to mentor 1 day early
  • Weekly meeting notes
  • Final presentation in D-term (poster or video as announced by the departments)
  • Term personal and team evaluations

Evaluation

You will receive an individual grade. There are tangible and intangible contributions to the MQP team. You will be assessed on your ability to:

  1. define and meet project and personal goals
  2. problem solve and use feedback
  3. work independently and collaboratively
  4. communicate
  5. synthesize information and make conclusions

Further, we want you to demonstrate:

  • leadership and teamwork
  • academic integrity

Important dates

  • Advisor meetings
    • A term: 3pm every Tuesday in Unity Hall 344
  • Mentor meetings
    • A term: 9:30am every Thursday in UH 341

Resources

Below you will find links to various resources that may aid in your MQP. We will periodically update this as the MQP progresses.

Relevant blog posts

Relevant videos

Tutorials

Tools

Code Resources

Datasets

Relevant papers

GAN Papers

Human Activity Recognition Papers


Tasks

Week 1

About

repo for GAN MQP projects

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 100.0%