A new Computer Vision Model (v2.5) including 1,454 new taxa

We released a new computer vision model today. It has 77,276 taxa up from 76,129. This new model (v2.5) was trained on data exported last month on June 18th and added 1,454 new taxa.

Taxa differences to previous model

The charts below summarize these new taxa using the same groupings we described in past release posts.

By category, most of these new taxa were insects and plants

Here are species level examples of new species added for each category:

Click on the links to see these taxa in the Explore page to see these samples rendered as species lists. Remember, to see if a particular species is included in the currently live computer vision model, you can look at the “About” section of its taxon page.

We couldn't do it without you

Thank you to everyone in the iNaturalist community who makes this work possible! Sometimes the computer vision suggestions feel like magic, but it’s truly not possible without people. None of this would work without the millions of people who have shared their observations and the knowledgeable experts who have added identifications.

In addition to adding observations and identifications, here are other ways you can help:

  • Share your Machine Learning knowledge: iNaturalist’s computer vision features wouldn’t be possible without learning from many colleagues in the machine learning community. If you have machine learning expertise, these are two great ways to help:
  • Participate in the annual iNaturalist challenges: Our collaborators Grant Van Horn and Oisin Mac Aodha continue to run machine learning challenges with iNaturalist data as part of the annual Computer Vision and Pattern Recognition conference. By participating you can help us all learn new techniques for improving these models.
  • Start building your own model with the iNaturalist data now: If you can’t wait for the next CVPR conference, thanks to the Amazon Open Data Program you can start downloading iNaturalist data to train your own models now. Please share with us what you’ve learned by contributing to iNaturalist on Github.
  • Donate to iNaturalist: For the rest of us, you can help by donating! Your donations help offset the substantial staff and infrastructure costs associated with training, evaluating, and deploying model updates. Thank you for your support!
Posted on July 20, 2023 05:39 PM by loarie loarie


Great to see some increasingly rare stuff in there! 4 new alliums just in this update! We're fast running out of North American alliums the CV doesn't know about!

Posted by wildskyflower 5 months ago

Have you thought about building an "adversarial" computer vision model? Instead of training on all Research Grade observations, train on all observations that are now research grade after user IDs overturned the original computer vision suggestion? I

f successful, this model could provide a 2nd opinion to the current model, or even suggest alternate IDs for existing non-research grade observations.

Posted by conorflynn 5 months ago

That's so good, hoping for a day which the flowering plants in the region of Minas Gerais mountains will be here on Inat, most of dipluson, Lavoisiera and some more

Posted by intelec 5 months ago

Aloe polyphylla added. With only 15 obs. Do we need less than the 100 photos, to be included now?

Lesotho endemic. Now 10 obs

And 87 cultivated - answers my question.

Posted by dianastuder 5 months ago

Great 14% of the plants were from southern Africa (not necessarily exclusively so) - with 10% of the worlds flora, we seem to be getting ahead of the curve! (No thanks to some help from the Spiral Aloe, which is a lovely surprize!)

Posted by tonyrebelo 5 months ago

Great! I have learnt to identify so many species thanks to AI suggestions. That's a wonderful tool!
Thanks to all who work on the CV!

Posted by bagli 5 months ago

Major thanks for all of these updates on the CV -- I enjoy filtering these by region. Thanks!

Posted by sambiology 5 months ago


Posted by kevinfaccenda 5 months ago

I really have noticed an impressive improvement in the algorithm over the last couple of years. It's still not perfect, of course, but it now can identify some really tricky species. And when it is wrong it's usually obvious it's wrong (like wrong kingdom, etc). What a neat tool. I kind of wish it would be run on all observations (maybe opt out or opt in) as an ID with a weight of zero (like it adds improving ID but doesn't count as research grade.) Or something. I also think it would be neat to make one that marks where on the photo the species was seen, as it sometimes sees several species. And maybe one that can collectively look at multiple photos. Ok i am getting ahead of myself, but i am just excited that it's been so great lately :)

Posted by charlie 5 months ago

@charlie in September 2014, less than 9 years ago, this xkcd comic joked that classifying whether a picture was of a bird at all using computer vision was a problem that would require "a research team and five years". At the time I was close to some academic computational neuroscience circles, and the reaction was basically "haha so true".

I think its pretty easy to underestimate how far inat computer vision has come.

Posted by wildskyflower 5 months ago

Monthly updates are motivation to check for Pending status, and retrieve more from the Needs ID bucket.
Hoping to see this in the next, or next, update.

Posted by dianastuder 5 months ago

Just a note that the first pie chart has a number error in the title - "1,4543 taxa ..."

Posted by ivanmunks 5 months ago

how to update? or is it automatic?

Posted by macro_expert 5 months ago

thanks for catching that typo - updated

Posted by loarie 4 months ago

August update coming soon?

Posted by dianastuder 4 months ago

Add a Comment

Sign In or Sign Up to add comments