I am working on importing our organization’s photos into Seals Wildbook. We work with harbor seals and would like to be able to use this database for the organization of our data as well as photo-ID. Before I do the many bulk imports that will be needed to upload all our photos I wanted to check a few things.
We do two-hour observations during which we take multiple photos of any seals that are spotted. One of our main goals for using Wildbook is to be able to see how many days in the past our marked seals were resighted. It seems to me that we could get this data from the “My individuals” page where it says the total number of encounters we have for each marked individual. My worry is that if we have multiple photos of a single marked individual for one observation and import those as separate encounters (i.e., rows in a spreadsheet) the total encounter value will be falsely inflated. In my mind there are two ways we can approach this issue. One way is to import just one photo from each observation day for each of the individual seals we saw (this would be the simplest thing to do). The second way would be to have those multiple photos of an individual per observation entered as a single encounter (i.e., use Encounter.mediaAsset1…). Does anyone know if one of those options would work better than the other?
Related to my first question, does uploading more images of our marked seals improve the ability of Wildbook to match future unidentified seals with the marked ones? If more is better than I lean towards the second option I proposed above even though it will take more time.
What you refer to as an observation is considered a “sighting” or “occurrence” in Wildbook. This represents a set time frame that the observation activity occurred.
One sighting is made up of several encounters, or instances where an animal was seen.
When you set up your bulk import spreadsheet, there’s an Encounter.occurrenceID field where you can assign the encounters in your observation a sighting/occurrence ID so that all of those encounters will be grouped under that ID. (list of bulk import fields and their descriptions)
Whether to include all photos of an individual during a sighting under a single encounter or upload each photo separately as its own encounter will be up to you. That could be a good question to post under the Community Research category of this forum to get feedback from other researchers about their preferred workflow.
Yes. When we look for opportunities to retrain our algorithms, we often turn to researchers with high confidence in their marked individuals. The most successful datasets are the ones with varied viewpoints of the same individual (left, right, front, etc.). The more examples we have of specific individuals, the more accurate your match candidates will be over time.
So it sounds like we should treat each of our two-hour observations as a “sighting” in Wildbook. I think this should work fine for our purposes, but I have a couple follow-up questions to make sure that is true.
One, is there a quick way to know the number of sightings a marked individual has been spotted during? As I said in my original post one of our main goals is to be able to see how many days/observations/sightings in the past our marked seals were spotted. On the “My individuals” page of Wildbook it says the total number of encounters, but I do not see anywhere that says the total number of sightings. Is that information elsewhere?
Two, if we were to use Enounter.occurenceID in our bulk import is there a way we can use Wildbook’s automatically created unique sighting IDs? It sounds like this will happen if you leave the field blank (says so in the list of bulk import fields and their descriptions). My worry with leaving the field blank is that Wildbook may treat each row/encounter as a unique sighting even if they should not be since I am not manually differentiating between sightings.
I’m still verifying if there’s a more direct way to get this info, but so far, doing an encounter search (Search > Encounter > Identity > Individual name) and the clicking on the Individual ID column in the search results will group together each individual with the associated Sighting ID next to it.
This is exactly what happens. If you don’t specify your own Sighting ID during the import, each uploaded encounter will be assigned its own unique one automatically. You may find it easier for tracking purposes to follow a standard format for your observations such as by date, time, and location separated by underscores (spaces don’t play nice in the Sighting ID field). This can always be manually edited after the fact if there’s a mistake, but it’ll be easier to get it right at the import level.
I verified that there isn’t a straightforward way to see this. This is because most animals are typically encountered once per sighting/occurrence so Wildbook will track the number of encounters associated with an individual. As an alternative, you can click on the export tab after you do an Encounter search and then set up the Excel sheet to sort the rows by the sighting/occurrence ID.
Thank you for verifying that. I have moved forward with that alternative and it is working. I would still like there to be an easier way to see the number of unique occurrences/sightings if possible. Would this be an appropriate feature request?