SMP with WS
I am excited about SMPs. They give me a chance to learn and get involved in something interesting outside of courses, and they are your opportunity to synthesize your learning into a larger project. An SMP can feel daunting at first, that is completely normal, and you are not expected to have it all figured out at the start. We will work it out together.
There are some projects that I would like to get started, and these are listed toward the end of this page. I am equally happy to supervise a different project of your choosing provided the sum of your points below is 5 or more.
- you can convince me of your genuine excitement about your project: 3 points
- your project is going to help people on campus, in the community, or online: 2 points
- your project is a learning opportunity for you or for me: 2 points
As your supervisor, I will not always have the subject expertise, and that is fine. In fact, it is our shared goal that during, and certainly by the end of, the project, you are more of an expert in the project domain than me. I will provide structure, accountability, and feedback on how you are thinking and communicating about your work. I will be available to you throughout — to talk through roadblocks, help you think clearly when things feel stuck, and point you toward the right people or resources whenever I can. We will work through the hard parts together.
An SMP is your chance to shine — to bring together everything you have learned and take ownership of something meaningful. The work is real and the expectations are real, but so is my support. What I ask of you is genuine investment: to be resourceful and responsible, and to be enterprising when you hit the inevitable obstacles. And obstacles do come ("If you are not failing, you are not trying.") As you know by now, problems are not a sign that something is wrong, they arise naturally when you take on something large and meaningful. Maintain open communication — we share the good times and the bad, and when you hit a wall, bring it to me and we will figure out a path forward.
This often involves finding relevant resources, doing the necessary research, learning the necessary theory, reaching out to relevant people, picking up the relevant software, developing helper tools, or wrangling data for the project. This kind of initiative is what senior-year work looks like. Most students find, in retrospect, that they were more capable of it than they initially believed. And facing problems, analyzing them, and systematically clawing your way out of them builds the kind of skill and resilience that will earn future supervisors' respect and help you succeed in industry or academia.
Projects
Below are a few projects that I have in mind. Some of these come from colleagues on campus who have a real need and are keen to collaborate. Please see me (Schaefer Hall 150) or contact me (wsaleem@smcm.edu) if you are interested in working on any of these.
College Data Analysis | Campus Living Laboratory | Graphics Pipeline | Machine Learning in Graphics
College Data Analysis
Problem: This project is to leverage existing AI platforms to analyze college data and derive actionable insights from it.
What is involved?: Colleges collect a great deal of data, e.g., on enrollment, retention, course outcomes, student demographics, and more. Automated analysis tools can surface patterns and support decision-making in ways that manual review cannot. You will work with Associate Dean Randy Larsen to understand his specific requirements and obtain the relevant data*. You will then identify and evaluate suitable tools, explore how they can be composed into an analysis pipeline, and build a dashboard that presents the results clearly and interactively. Your final deliverable will be a dashboard that satisfies Dr. Larsen's data and interface requirements as well as any other technical requirements.
* - for confidentiality purposes, you will be provided synthetic data.
Prerequisites: A general computer science background suffices. Many tools will be learned as the project proceeds.
Campus as Living Laboratory Dashboard: Monitoring Sustainability Data
Problem: This project is to build a dashboard to monitor and analyze campus sustainability and research data.
What is involved?: There are many environmental factors to monitor at St. Mary’s College of Maryland, e.g., at the Kate Farm soil temperature, moisture content, pH, salinity, nutrient levels (nitrogen, phosphorus, potassium), and organic matter content. The ENST program has and other College offices oversee or have installed sensors that generate data across campus. They want to build a dashboard that interactively visualizes and analyzes the data. You will work with Dr. Barry Muchnick, the Director of the Physical Plant, and the Sustainability Fellow to learn about available data and ongoing and emerging research to their data requirements. Your deliverable will be a data visualization and analysis dashboard that satisfies the ENST requirements as well as any other technical requirements.
Prerequisites: A general computer science background suffices. A basic understanding of hardware is preferred. Other tools will be learned as the project proceeds.
Machine Learning and Signed Distance Functions for 3D shape representation
Problem: This project is to explore the use of machine learning for 3D shape representation and rendering.
What is involved?: Signed distance functions (SDF) for 3D shape representation are an emerging area of research. Recent approaches use machine learning to develop the SDF. You will learn how SDF can represent shapes, how machine learning is used to learn SDF, and propose and develop a reasonable application.
This project will require research, math, and implementation, especially graphics programming. Learning the required skills is part of the project.
Prerequisites: This project is math- and programming-heavy, and will require comfort with both. Background in 3D graphics, e.g. from COSC 338 Computer Graphics, is helpful but not strictly required — the willingness to learn is more important. Courses in linear algebra and calculus will be relevant. Comfort reading and engaging with technical and research literature is essential.
References:
- Neural Geometric Level of Detail: Real-time Rendering with Implicit 3D Shapes, Takikawa et al., CVPR 2021.
- Advances in Neural Rendering, Tewari et al., ACM SIGGRAPH 2021.
Simulating the graphics pipeline
Problem: This project is to build a software simulation of the graphics pipeline.
What is involved?: We take for granted the operations performed by the graphics pipeline which is hard-coded in the GPU in our computer. This project will simulate the stages of the pipeline in software with the goal of developing a deeper understanding and exploring modifications and extensions to the pipeline.
Prerequisites: Familiarity with the graphics pipeline, e.g. as gained in the course COSC 338 Computer Graphics, is required. A good grip on mathematics and programming, and comfort reading technical content is also required.
If you have made it this far, great! I wish you all the best for your SMP.
alles Gute und viel Glück!