Implicit Surfaces & Independent Research

Video Description

Implicit surfaces are a fascinating microcosm of computer science. On the surface, they're simple: define a 3D solid as existing anywhere a function ℝ³ → ℝ is negative. This simplicity makes them a compelling backend for solid modeling and computer-aided design: they're compact, expressive, can be evaluated at arbitrary resolution, and support easy constructive solid geometry (e.g. union and intersection). It's possible to write a renderer in a few dozen lines of code, but this simplicity hides a deep rabbit hole. In this talk, we'll dive head-first down that rabbit hole, exploring optimized rendering and strategies for fast evaluation. We'll visit many fields of computer science along the way: graphics algorithms, data structures, numerical computing, and even compilers! The talk will also discuss the nature of independent research: I've been exploring this field as a hobby for over a decade post-graduation, and have published work at a top venue as an independent researcher. This is an uncommon path, and I'll share what I've learned along the way. -------- Speaker intro: 0:10 Talk begins: 1:22 This is a recording of a talk given at the Harvey Mudd College CS Colloquium on March 7, 2025. Thanks again to HMC for inviting me to speak with their students!