World Flora Online and iNaturalist Partnership?

I happened to run into a 2023 paper, The big four of plant taxonomy – a comparison of global checklists of vascular plant names by David Costa et al. (https://doi.org/10.1111/nph.18961), today, and after reading it decided to check up on World Flora Online again. I discovered it about a year ago but it seemed then it like wasn’t quite “mature” yet.

Anyways, I was browsing around the site when I noticed they have their meeting notes posted (https://about.worldfloraonline.org/documents). In the most recent set of notes (https://about.worldfloraonline.org/images/uploads/documents/21st-WFO-Council-Final-Report-November-2023-virtual.pdf) there’s a section from pages 11 through 15 about possibly adding identification resources to WFO.

In their slide deck for that section, is a slide which says:

Image recognition: The WFO Council is recommended to organize a
partnership with Pl@ntNet, to try and realize a strong link between both
facilities, pointing users to this external option for identification of plant
images.

[Pl@ntNet seems the only organization with a strong interest in such a
collaboration. WFO can also stimulate Consortium members to provide
reliably identified photographs, further improving the AI routines.
Identification using digital images of herbarium material to be investigated
further.]

Does anyone know what that comment is about? Just curious if this is something that was entertained and then decided against? I couldn’t find any mention of it in the forum (though maybe it isn’t the type of thing that would have been). Seems like this kind of partnership between WFO and iNaturalist would be pretty awesome!

8 Likes

I know in past discussions, iNat has been weary to open up CV abilities via an API and it seems like they mostly want to keep the CV only for photos which are actually uploaded to iNat. I’m not sure what the reasoning for this is, it seems like it could be a great revenue stream to charge access to the API.

I do know that the Pl@ntnet CV is not as good as iNat, at least in Hawai’i where it was 20-30% less accurate. I’ve used pl@ntnet before and the lack of editability really bothers me. They contribute all these bad IDs to GBIF, but then none of them are reviewed by people

7 Likes

Huh, that’s interesting. I could see an argument being made that they want an free and open API and don’t want to charge access for it; but outright not wanting to expose the functionality at all is a bit curious. I’d like to hear more about how that decision was come to.

Is the CV functionality not where the monetary value of iNat lies?

2 Likes

In the past (5-7 years ago) i thouight botanical gardens and ilands were participating with Plantnet and got their own local version of Plantnet https://plantnet.org/en/microproject/
As they only need Plants certainly would use a dedicated plant app.

I thought Naturalis had an api (Naturalis Identification API, NIA) which one could use 3times a day without an account https://docs.biodiversitydata.nl/en/latest/

I think this link is better

https://multi-source.docs.biodiversityanalysis.eu/

https://multi-source.identify.biodiversityanalysis.eu/

Till my surprise iNaturalist is included, this explains a remark i read somewhere from one othe programmers/developers, i thought it was fully depending on observation.org and waarneming.nl photos.
https://multi-source.docs.biodiversityanalysis.eu/contributing-data/images/

1 Like