Thank you for your reply and for looking into it.
This one could be a candidate too?
Which ID?
wow this is another one of those cool urls that I have no idea that it already exist.
i also learned only today of another such unlinked inat URL - https://forum.inaturalist.org/t/interpolating-coordinates/5170/
The first (obviously wrong one) seems to be created by wrong suggestions of the app. I can reproduce the wrong suggestions.
the subsequent IDs (‘Parthenocissus’) are the correct ones.
If it was chosen directly from a computer vision suggestion then the ID would have the computer vision icon.
The user manually entered it, as far as I can tell. They could have chosen a suggestion from computer vision then manually typed it in, but I’d be surprised if that was the case since it’s not near the top of the suggestions.
My mistake, if the ID was added via the Android app, it doesn’t always mark the ID as such.
Hm, you are much better in iNat behaviour than me. But this sign I get in all cases I add an ID to an observation. Regardless weather I chose the computer vision models suggestions, look to the CV suggestion before I type a species name myself or only type a broad ID (like "Magnoliopsida) where I don’t have to look at the CV’s suggestion.
Anyway, the result here in my example I asked for, was the same (the computer vision suggested “Hedera”, “Hedera” was entered by the user, but it is wrong). If the wrong suggestion is intended here, the computer vision model must be trained better, before it can recognize this picture correctly or any other reason, this must be decided by iNaturalist. For me, it’s not, what I would expect as a user, so I assume it’s a fault.
Thank you very much for having taken a look at it.
Just to note the hexagon/expected nearby problem has been raised … again. https://forum.inaturalist.org/t/geomodel-hexagons-still-lacking-basic-error-checking
I amended my comment, it could have been added via CV.
If I upload that photo in the web uploader and use the exact same coordinates of the observation, I see this:
I think it’s working correctly. It’s showing Parthenocissus as the “pretty sure” genus and Parthenocissus tricuspidata as the top suggestion. I can see why it would also suggest Hedera helix, the photo is from pretty far away and probably looks like other Hedera photos.
Thank you so much for your thorough testing of the issue.
so hopefully it will be more the quality of the picture (angle, light, distance from the plant, …) than the recognition quality of CV. I sometimes get really strange results during field work, but there is the time missing to do longer technical tests. Than I keep it “cf” for studying the literature at home.
For critical to separate species, the technical assistance of the app continue to have its limitations. But perhaps this is not the goal, too.
One more question about this…
I am wondering if iNat wants to know about common plants in a location falling outside of hexagons or not?
One today, the only suggestion in the list was oxeye daisy (so the person chose that). The plant was Cakile edentula. I saw that if I chose “not expected nearby,” then C. edentula was at the top of the list.
Should we be making a notification somewhere when something common and expected has fallen outside of a hexagon to help with future updates to the GeoModel (hope that is the correct language)? I just want make certain I’m using the system correctly. Thanks!