Preconditioning in NeRF; today's reading.

deep-learning preconditioning today-i-read

  1. Keunhong Park, Philipp Henzler, Ben Mildenhall, Jonathan T. Barron, and Ricardo Martin-Brualla, “CamP: Camera Preconditioning for Neural Radiance Fields,” ACM Trans. Graph., 2023.

    tl;dr: NeRF joint optimization of camera parameters and scene reconstruction. Uses a left preconditioner for each camera’s parameters (Zero Component Analysis (ZCA) whitening transform (Kessy et al. 2018)), derived from a projection function; apply this at the initial iteration of the optimization. The new method is implemented on top of Zip-NeRF (Barron et al. 2023).

© Amy Tabb 2018 - 2023. All rights reserved. The contents of this site reflect my personal perspectives and not those of any other entity.