Deep Genera)ve Model Course Project Guideline
Project Overview
Objec&ve: Integrate deep genera,ve model methods into your ongoing research projects. This
project aims to bridge the gap between theore,cal knowledge and prac,cal applica,on,
allowing students to explore the poten,al of genera,ve models in their research domains.
Steps:
- Iden,fy a challenge or aspect of your current research where genera,ve models can be
applied.
- Implement and integrate a genera,ve model solu,on.
- Analyze and compare the results with exis,ng methods in your research.
Code: While leveraging exis,ng libraries and tools is encouraged, students should ensure
significant code development on their part, especially in integra,ng genera,ve models into their
research context.
Milestones:
- Research Integra,on Proposal (Due 11/12): Outline how you plan to incorporate
genera,ve models into your research.
- Final Report (Due 12/14): Present the final results and analysis.
Report Language: Korean or English
Team: You may form a team of up to two people. If you choose to work in a team, it is expected
that the project will reflect double the workload compared to a single-person project. If you
choose to work in a team, the peer review will assess the contribu,ons of each member to
specific parts of the project.
For Students Already Working on Genera&ve Models: If your primary research already revolves
around genera,ve models, this project presents an opportunity to further deepen your
exper,se. Consider the following approaches
- Advanced Explora,on: Dive deeper into more advanced or recent techniques within
genera,ve models that you haven't explored yet.
- Compara,ve Study: Compare mul,ple genera,ve model techniques to determine their
strengths, weaknesses, and best applica,on scenarios.
- Hybrid Models: Inves,gate the combina,on of genera,ve models with other machine
learning or deep learning techniques to create hybrid models.
Details on Proposal (Due 11/12):
Your project proposal (2-3 pages) should include the following, in this order.
1. Names of all team members, date, and project ,tle
2. Introduc,on:
a. Your research project overview and its current status.
3. Iden,fied challenges or aspects where genera,ve models can be applied.
4. Proposed genera,ve model solu,on and its expected impact.
5. Bibliography of related works in genera,ve models.
a. Make a high-quality bibliography where each reference has the complete list of
authors, ,tle, the complete cita,on, etc., and a URL or DOI.
6. Submission
a. I-Class: One team member uploads the proposal in pdf format.
Details on Final Report (Due 12/14):
The final report should be wriaen for anyone new to the topic. The descrip,on should be
sufficiently detailed that anyone could repeat what you have done. No code should be included
with the report itself; instead, the algorithms should be described with mathema,cs and/or
with pseudo-code, like in a journal paper. A typical report might be 6-8 pages plus figures,
tables, and references, but there is no page limit.
Your final report should include the following, in this order.
1. Recap of the research project and the iden,fied challenge.
2. Detailed method of the genera,ve model solu,on.
3. Results & Discussion, comparing the genera,ve model's outcomes with exis,ng
methods.
4. Cri,cal Analysis of the integra,on process.
5. Conclusion & Future Work, including poten,al improvements or other genera,ve
models to explore.
6. References.
7. Contribu,ons: If it is a team project, a detailed breakdown of who did what.
8. Submission:
a. I-Class: One team member uploads the final report (in pdf format) and code (in
zip format)
Evalua7on Criteria:
1. Understanding of Genera,ve Models: Clear presenta,on and understanding of the
chosen genera,ve model's mechanics and ra,onale.
2. Integra,on Quality: How seamlessly and effec,vely the genera,ve model is integrated
into the research project.
3. Result Analysis: Depth of comparison between the genera,ve model's results and
exis,ng methods.
4. Implementa,on and Code Quality: Organiza,on, documenta,on, and overall quality of
the code.
5. Report Quality: Clarity, structure, and use of visual aids in the report.
6. Cri,cal Analysis: Discussion of challenges, solu,ons, and any devia,ons or unexpected
outcomes.
7. Timeline Adherence: Mee,ng specified milestones and deadlines throughout the
project.