Flagger | Content Author | Content | Reason | Flag Created | Resolved by | Resolution |
---|---|---|---|---|---|---|
ny_wetlander | Genus Lordotus |
many misidentification due to erroneous iNat autosuggestions |
Sep. 18, 2021 16:31:24 +0000 | matthewvosper |
Can only be resolved by correct IDs |
I agree that bombyliid IDs are pretty scary on iNat by and large, but I don't think there's anything that can be fixed by a flag, though. (Flags don't control CV suggestions.) The solution is just to better ID the misidentified observations so there's fewer false positives in the algorithm's training set and so the species won't show up as a "seen nearby" option in places where it doesn't belong (like out east). If you want some crowdsourced help bumping things up, you could post this on the wiki for problem taxa like this: https://forum.inaturalist.org/t/computer-vision-clean-up-wiki/7281
@tristanmcknight, I created a comment so I would have something to link to when I added it to that forum post :)
CV AI regularly and incorrectly defaults to this genus in North America where there is a fuzzy yellow fly, but the genus only occurs in western North America. It would be best if the "suggested" ID were to the subfamily level Bombyliinae. For example, right now of the first 44 genus level IDs, 6 are incorrect, including one that is in a different subfamily.
A clear photo of the wing venation is very helpful, if not necessary for correctly identifying this genus.