Add more data scenes

Thank you for your great work! I want to use the co3d datasets to do research, but the scenes are not enough for me. I need to prepare some special scenes. So could you share that how can I annotate my own images with depths and the camera intrinsics and camera poses that align with your dataset? Sincerely hope for your help.

Start by reviewing the repository's dataset tooling and any existing guidance for CO3D scenes. Trace how depth, camera intrinsics, and camera poses are represented, then document a reproducible path for annotating custom images that aligns with the dataset. Done means a newcomer can follow the documented process without needing an unanswered issue thread.

Written by the indexing model from the issue text.