What is the entire URL out of the browser, exactly where the error occurred? Flukebook | Login
Can you describe what the issue is you’re experiencing?
I have entered a new encounter of fin whale.
The workflow state is unproved, I would like to ask if it is possible to approve such encounter.
Can you provide steps on how to reproduce what you’re experiencing?
In the photo gallery of the same whale, I did several matches with different algorithms, and got some results.
What can I do for each search result to know if this whale has already been identified or not?.
You can change the approval state by clicking on the edit button in the Metadata section.
It’s up to you to review the match results on the Encounter to verify whether any of the candidates are the same whale. You can either confirm a match on the page by clicking the checkbox next to the result that looks correct, or you can set it manually by entering the Individual ID on the Identity section of the Encounter.
Each image in this Encounter has its own set of match results, so I’m not sure which one you’re referring to. Can you link to the match result page you’re looking at or post a screenshot so I can take a look?
What’s happening here is that the annotation from the match candidate on the right isn’t placed correctly. See how it leaves out part of the fin on the left? I suspect that and the low quality of that particular image makes it harder for the detector to determine where the fin starts and the water begins.
You can try going to the encounter that image came from, removing the annotation, and then adding a manual annotation again.
The numbers are match scores. The higher the score, the higher the likelihood of a match. What the score represents is slightly different for each algorithm. If you hover your mouse over one of the results, you’ll see an “inspect” button appear. Clicking on it shows a side-by-side comparison of the images along with a visualization of the areas identified by the algorithm as a match (as in your last screenshot).
@jason for visibility in case you have any details to add about how the match scores are determined.
I need more time to find out the specifics of how each algorithm calculates its match scores.
Those numbers next to the scores (700723 and c3f194ab-d4b7-4) are the names of the whales yours matches with. If you click on them, they open links to their Marked Individual page.
Thanks for your patience. I asked Jason for more insight on Flukebook’s match scores and this is what he shared:
All three algorithms have relatively inscrutable scores. They’re not probabilities and not especially meaningful but are rather distance metrics of one form or another and will vary from trained instance of the algorithm to another trained instance.
PIE v2 can sometimes show a significant score gap between a strong match and a weak match. But the scores themselves are not meaningful.
The rankings are what you see on your match results page. 1 represents the match with the highest confidence and the rest of the matches are displayed in descending order of confidence.
