Noisy LLFF Synthetic

The dataset used in the paper is a large-scale paired dataset for training a NeRF-agnostic restorer. It consists of 8 virtual scenes, each with 400 images.

BibTex: