Automated validation of match results

I will be quantifying the match success of flukebook for the killer whale population I am working with by uploading photos of known animals (with their IDs specified) to create a catalog to match against, and then uploading additional photos of known animals (without their IDs specified) to see where the correct whale is ranked in the match results. Is there a way that the match results can be exported in bulk to assess the suggested matches against my known IDs for those photos and to then bulk update the metadata for those photos with the correct IDs once the matching success has been assessed?

For example, an ideal workflow I would picture is:

  1. Submit reference and test photos as described above.
  2. For all submitted photos put through the matching algorithms, export a table of IDs of possible match ID by their rank (for each algorithm)
  3. Upload an updated metadata sheet that includes the correct IDs for the test photos.

Is anything like this possible? Or would I need to manually review the possible matches for each photo, manually record the correct ID’s rank for each algorithm, and manually select the correct match (if found)?

On a somewhat related note: is there documentation on how to export ID results once matching has been completed for a series of encounters?

You would be able to export Search results in Flukebook but that would be based only on the encounter data, and not the match results. It’s an interesting use case, though.

Tagging @jason for visibility.