PlantNet Revised

PlantNet’s offerings have been revised. It now gives a score (a fraction, to two decimal places).

The release isn’t announced on the home page.

It says that it is reporting from North West European species, but it’s still offering North American trees, possibly because they’re in PlantNet’s database from cultivated or erroneously identified specimens.

It will be interesting to see how the updated AI performs.

It is not announced as we are still working on it to bring in some of the suggestions that the community has made. Some things may be possible and others not, it is difficult or impossible to do this testing on the test site.

Well it worked for you for the African tree, but I did notice that they had a dedicated data base of African trees.

Then I found this interesting, since I knew the name, Alphonse Lavallée

We had friends who produced export grapes on their farm in Paarl (Western Cape) and they always talked of this cultivar with pride.
Another cultivar was haanepoort which is still a favourite table grape in Cape Town. Wish we could buy them in London.

Origin : NorthAfrica, in the time of the ancient Egyptians, who used it for making wine.

How long between posting an observation can we expect a PlantNet ‘suggestion’?
I was expecting one for this post Geranium lucidum; using Stace | Observation | UK and Ireland | iSpot Nature
and I have already drafted my response, which is about how good PlantNet has been on Geranium lucidum with some comments on why this might be. The current comments on the post are about the distribution and frequency of G. lucidum, which my PlantNet comments also addresses.

But as yet no sign of PlantNet on the post.

I thought this example of Plantnet getting it wrong was mildly amusing.
It went all tropical when it was shown a nettle that had tried to press itself like we used to do with plants when cameras had expensive film in them.

the link.

Joc, none of the 7 UKleaves contributions by me since May 19th have been commented on by P@Net. Some, if not all, are pretty obviously plants!

We await further elucidation from the centre…