--chunk-size: The number of consecutive frames to use to robustify the relocalisation.
camera_poses.txt: File containing a list of camera poses (they have to be accurate frame-to-frame, the origin doesn't matter for this specific script) in the same format as the relocalisation poses.
Usage:
./src/compute_batched_relocalisations.py [--chunk-size N] <camera_poses.txt> <reloc_poses.txt> <out_folder>Params:
--chunk-size: The number of consecutive frames to use to robustify the relocalisation.camera_poses.txt: File containing a list of camera poses (they have to be accurate frame-to-frame, the origin doesn't matter for this specific script) in the same format as the relocalisation poses.reloc_poses.txt: File containing a list of relocalisation poses, format as described here: https://github.com/WaldJohannaU/RIO10#evaluationout_folder: Folder that will contain the output relocalisations. Same format as above.