Not sure if this is exactly a bug, but the auto-ID sometimes suggests monotypic genera. For example, this observation: https://www.inaturalist.org/observations/24639253 suggests the genus Sanguinaria. There is only one species of Sanguinaria. This seems to imply that it is more confident in the genu…

We discussed this and we’re pretty comfortable with it if the genus is marked complete. So it would affect most vertebrates and some other clades.

[image] jdmore: If an identifier doesn’t happen to be aware that the genus is monospecific, however, and decides to be prudent and choose genus, then that still requires an extra ID to move it along to species level later. There are also identifiers who are aware of taxonomic revisions, and t…

A lot of plant genera in particular probably need updating then to account for this. I don’t think many of those are marked “complete”, that tends to come with bird or mammal taxonomy more often.

This really does need to be addressed. The taxon that comes to mind for me is Malosma laurina, the only member of a SoCal genus Malosma. I can’t tell you how many times I’ve IDed genus-level Malosma observations to the single species, and then had to write out a comment saying “this is actually the …

Probably just a matter of time before the single polytypic and wide-ranging species in Diadophis gets split.

Wait if this is true then this issue shouldn’t be happening… [image] New Vision Model Training Started So genera are booted from the model as soon as a single child species is separately included in the model?

They’re in the “We’re pretty sure this is in the genus” section of the drop-down (which is programmatic/not a “cv-trained” taxon), but they’re not included separate as taxa that are listed below that part.

[image] grmorrison: I can’t tell you how many times I’ve IDed genus-level Malosma observations to the single species, and then had to write out a comment saying “this is actually the only species in this genus, so no need to ID to just genus level”. A lot of us can relate. One of these days, …

Haha, yes, there are actually plenty of observations at “Ginkgoopsida” level on iNat. Most fly under the radar because of being casual. The monotypic genera I most commonly bump to species are Galax and Medeola.

I don’t really get the point of this for genera unless the CV is 100% accurate at monotypic genera, because if someone blindly clicks ‘we’re pretty sure’ on species it is the CV directly contributing to becoming RG whereas it just makes it easier for human IDers to find if they click the ‘we’re pret…

Don't suggest genus for monotypic genera (computer vision)

Feature Requests

upupa-epops November 6, 2021, 9:29pm 22

Wait if this is true then this issue shouldn’t be happening…

Topic		Replies	Views
Change computer vision suggestions to only above species level Feature Requests	31	3085	March 26, 2019
Automatically identify monotypic taxa as the subordinate taxon Feature Requests web , search , android-app , taxonomy , mobile-app	40	4048	September 15, 2020
Computer suggestion being ''too precise'' General	22	1338	January 24, 2024
Computer Vision should take into account fraction identified to species General	16	797	January 9, 2025
Force computer vision to back off on the specificity of suggested IDs in regions with cryptic or hard-to-identify species Feature Requests identification , computer-vision	94	3209	July 20, 2026

Don't suggest genus for monotypic genera (computer vision)

Related topics