Is your feature request related to a problem? Please describe.
Currently we need to manually supply coco_image_filenames to annotations_to_coco to set the file_name value in the COCO json output.

Describe the solution you'd like
For framelabels.json the file_name should be:

f"sub-{subject_id}_ses-{session_id}_cam-{cam_id}_frame-{frame_number}.{frame_file_ext}"

For cliplabels.json:

f"sub-{subject_id}_ses-{session_id}_cam-{cam_id}_frame-{frame_number}"

We could have a helper function that takes input_path, subject_id, session_id, cam_id to generate this list of filenames to pass to annotations_to_coco.