Can you describe what the issue is you’re experiencing?
A user has uploaded many photos to have catalog images of some lynxs in Wildbook. However, in doing so, they have uploaded photos with multiple lynxes, and as a result, in the given encounter, there are green boxes around lynxes that are not the ones identified. (For example, in a photo with 3 lynxes, the green box is around the mother, while the identified lynx is the cub).
What would be the best way to fix this?
Remove annotations from this encounter and perform the ID in a clone encounter?
Remove the image with more than one lynx and resubmit the cropped image.
How do we proceed when there are more than one lynx in the pictures?
If this is a bulk import report, send the spreadsheet to services@wildme.org with the email subject line matching your bug report
It looks like this user decided to upload all of their lynx sightings on a single encounter as the camera trap dates are across different dates (and likely of different animals). You may want to reach out to them to ask them to resubmit one photo at a time for their encounters since an encounter represents a specific date and time that you came across an animal (or its friends) and delete this encounter and its clones after they do that.
For single images with multiple lynxes, like the one with the mother and cubs you mentioned, Wildbook already accounts for this. It automatically clones the encounters of any additional body annotations it detects. So for example, in the first photo where the mother appears, her annotation is the primary one and if you click the arrows below to highlight the other annotations, you’ll see a link to Go To Highlighted Encounter appear that will take you to the cloned encounter for the cub where you can manage its ID from there.
Thanks for the quick answer, Anastasia,
The user wanted to upload images from the Lynx called Terra. The main goal was to have images from Terra to compare with. Because of this, no GPS location or date is relevant.
Is it better to do a bulk import than just submit an encounter in this way? Or in what way?
Thanks in advance
Thanks for your patience. Their best bet to build a catalog of Terra’s image is to create encounters with one image each. That can either be with individual uploads or as a bulk import with one media asset per line.
Encounters were meant to document one animal at a specific place and time. Typically, when you have multiple images of that animal in a group on one encounter, it gets more complicated to track which animal is the right one across group photos because you can’t choose which is the primary annotation representing that animal unless you manually annotate it.