3D Reconstruction from Sparse Data
I love processing interesting data so that it can be viewed three dimensionally. Part of the fun for me is the challenge of deciding which workflows to use (ie. LIDAR, point clouds, photogrammetry, neural radiance fields, gaussian splats, and others).
The Slate Auto company is of particular interest to me, I love their DIY approach and have been excited about the release of their new vehicle. I reverse-engineered a 3D model of a custom designed Slate using their Slatemaker car customization tool and RealityKit on macOS. I then placed the 3D model in my driveway.
Several years ago I created a video that describes how to use a photogrammetry workflow to generate a model from stereoscopic images taken on Mars by the NASA Curiosity Rover during Sol 1000.
I get a kick out of the fact that I still occasionally receive emails from students and teachers that are using the video in their classes.