Stonehenge

The original Stonehenge video is here: https://www.youtube.com/watch?v=UrLUfDhHe-A 1) I use the yt-dlp package for downloading the youtube video (1440p) 2) used ffmpeg to extract frames at 1 fps, totaling 534 frames. 3) used colmap 3.13dev0 to create poses for the images. It uses vocab_tree_matcher and detects 378 images. Finally I undistort it with colmap image_undistorter 4) I uses splatfacto-big with bilateral-grid for 100k iters. Allows all images to be training dataset (no eval dataset). 5) imported the gaussian splat to supersplat and publish it with SOG