I am sure this has been asked before. I tried searching. I really did. How do you handle providing an ID for a plant when it’s not the most prominent one in the photo? There is a plant in Colorado that I identify a lot. It has a prominent orange flower so the computer model picks it up easily. But sometimes while the flower stands out the plant that you see the most of in the picture is not the one with the orange flower- it’s something else entirely. How do you handle this? I see three options:
Validate the ID even though it’s mostly a picture of something else
Leave it unresolved
Ask for more/better pictures or to crop the picture
If someone can point me to a thread where this has already been discussed happy to go read it instead. Thank you!
I believe that falls under the category of respecting the observer’s intention about the focus of the observation. If they have identified it as that orange flower, it’s fine to confirm the ID even if another plant takes up most of the photo.
If the observer has left it as “unknown” or “plants” it is still fairly reasonable to assume the showy flower is the intended focus, but if you’re slightly uncomfortable making that call you could note “for the orange flower” and be willing to withdraw in the unlikely event they come back and say the observation was meant for one of the other plants.
If the observer has correctly IDed the orange flower, I will agree with it, and suggest that they also provide a cropped photo (after checking that they’re still active)
If they’re still active, but didn’t ID it at, I will comment something like “there are several species here, I’m not sure which one you hope to identify. The orange flower is species X, but there are also species Y and Z. Could you clarify, or maybe add a cropped photo?”
That might seem like a lot of effort, but I like to think that in at least some instances I am educating a user (especially if they’re new) that the subject of a photo isn’t always obvious.
i ID a lot of western/clark’s grebes, and they look fairly similar and tend to form mixed rafts, so i will often run into an observation of a raft containing both species with a single species ID chosen from the computer vision.
In cases like this I will respect the observer’s (presumed?) wishes and agree with that ID, but also leave a note like “ID for the bird in the back left. the front center one is ____” or something. i do this both in case the user meant to pick a different subject and got confused by a CV suggestion but also just to prevent confusion to anyone else who might look at that observation in the future.
I have ID’ed an organism with the note as to which part of the image was being ID’ed. In one case, the original poster commented a different organism was wanted, but thanked for what had been ID’ed.
In which case the logical thing for the observor to do would be to create a second observation to capture the second species that can be identified and add notes to that effect to both observations.
Nearest I had to that myself was when most pictures in a series only had a Great Tit in them but one image had a less obvious Blue Tit in it as well. I uploaded the same picture to two different observations and someone jumped in and disagreed with my Blue Tit ID with Great Tit. Quick note in the comments sorted that out.