Open Positions
PhD students
Our lab will be hiring several new PhD students this cycle (to start in September 2027), to work on problems related to the interaction of geometry and the real world (see our Publications for a sample of the research you could work on as part of our lab).
One of these new PhD students will lead our lab’s efforts on designing algorithms that can detect firearms during the 3D printing process. Given the danger of 3D-printed firearms (which can be undetectable and unregistered, and are increasingly used in crime), several states (like California and New York) have enacted laws requiring 3D printers to incorporate software that detects whether the printer is being asked to print a firearm. However, existing methods are either quite easy to fool or require revamping the entire 3D printing pipeline around closed-source tools. Building from our prototype firearm-detection benchmark, this project will build a new state of the art for this task. Many project-sized questions arise from this:
- At which stage of the 3D printing pipeline should detection be carried out? Slicers have access to meshes, while printers only see machine instructions. Slicers can (maybe) access the internet or bigger memory; printers can’t. But printers know the object’s true scale…
- How can one validate that a set of machine instructions corresponds to a given mesh? Any slicer-level detection will need this kind of authentication.
- Firearms are often printed in parts: how can one detect which components can only belong to a firearm, and which ones have benign uses?
- …
While this PhD will begin by answering these research questions in collaboration with our partners in academia, government and the non-profit sector, it will grow into broader questions about how 3D geometry processing can help promote safety and reliability in manufacturing.
How to apply
If you are interested, please apply to the Columbia Computer Science PhD program and list Silvia Sellán as your preferred advisor. In your statement of purpose, you should mention your specific interest in the project described above. If you want to make your application stand out, you may (but are not required to) familiarize yourself with our detection benchmark and use your statement to mention potential detectors and adversaries you could contribute. Please refrain from using generative AI tools for this task, as you will be asked about it in depth if selected for an interview.
If there is any context you would like to give your application that does not fit in any of the application’s pre-set fields, feel free to contact Silvia via email. We particularly encourage students from non-traditional backgrounds to apply: we will consider your application in its entirety, and do not rely solely on your number of prior publications or the ranking of your undergraduate university to make admissions decisions. We are a diverse and growing lab, and we encourage students of every gender to apply. We also do not take a person’s nationality or likelihood of obtaining a visa into account when making admissions decisions.