Another example here
https://wildbook.seadragonsearch.org/encounters/encounter.jsp?number=d6e4cd3a-34c8-4b2f-8c79-143a51e352cc
Once WA image was detached, righted and re-attached, a subsequent manually started search only finds one match, from TAS… why? it shouldnt be searching against these…
https://wildbook.seadragonsearch.org/iaResults.jsp?taskId=2e6cb011-99f8-431a-abf9-16ad16768ca0
Hi @tanyastere does this have a ticket yet? I also have another example here. I reattached an image, and the new right head search compared to other states, which shouldnt happen…
https://wildbook.seadragonsearch.org/encounters/encounter.jsp?number=1fdada15-7851-4342-be90-ce80e2836ed1
thanks, Nerida
Hi Nerida,
Sorry we missed this. I’m currently having a look and will post an update before end of day.
Thanks.
Hello again.
Do you know which route was used for these ID jobs? I can see in the log for the task that location was not used to compile the matching set. Running them again from the encounter page succeeded however. Both encounters now have results specific to their location.
Was this the automatic ID run from an encounter submission or was it initiated from another page?
Is there a possibility that the location ID for the encounter you wished to match was set after the initial submission?
Thanks for this additional information and for your patience. If you have seen other examples I’d like to have a look at them too. I think that any affected will share the same ‘route’ to the ID pipeline. I’ll assign a ticket when we we can glean a clearer idea of this behavior’s source from your response.
I believe I see the behavior you are describing now. New assets added to the encounter after it’s initial creation has completed are being sent to detection and ID, but the ID portion fails to take into account location.
I have a fix in for this under WB-1526. Please let me know if your problem is resolved.
seems to be, thanks Colin.
1 Like