A new Computer Vision Model (v2.6) including 1,399 new taxa

We released a new computer vision model today. It has 78,387 taxa up from 77,276. This new model (v2.6) was trained on data exported last month on July 16th and added 1,399 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 August 23, 2023 08:32 PM by loarie loarie


we love the new updates, i got a question tho, what does it mean lost (23)? it means they were in the database but dont exist anymore?

Posted by intelec 11 months ago

74 new plants for Africa. Thank you!

Posted by dianastuder 11 months ago

the lost species dropped below the obs threshold required to be included in the model, usually its because of the community changing its mind and adding new identifications, e.g. the Encrusting Rose Sponge example here https://www.inaturalist.org/blog/69958-a-new-computer-vision-model-including-4-717-new-taxa

Posted by loarie 11 months ago

It would be nice to have more species from the neotropics represented in the model.

Posted by deniszabin 11 months ago

@deniszabin: The great thing about the iNat CV model is that inclusion is entirely dependent on there being enough iNat observations and photos. If you can help identify observations so that there are 100+ photos of a particular species, you can expect it to appear in a future version of the CV model: https://forum.inaturalist.org/t/how-are-photos-selected-for-cv-training/42403/11

Posted by rupertclayton 11 months ago

It's so cool that our observations are being used to train the model on new species!

I have a question... Do casual observation photos such as photos of a rare or endangered animal within a zoo ever get used for the training model?

Posted by jarronevsbaru 11 months ago
Posted by thebeachcomber 11 months ago

@deniszabin choose your taxon - check the About tab, Pending or Included?
That 100 photos translates to about 60 obs.
For your chosen taxon. If you can help retrieve obs that may be languishing in a higher taxon level. Or you can deliberately add a few more of your own (the model chooses from various observers so not 100 of YOUR pictures)

There are good months where I can see my Fynbos Rambles have helped to add That Species!
Since the date when a taxon is added to CV is not available - beyond the links in each blog post - I leave a comments on my obs.


And motivating nudges on obs where the taxon is closing in on CV
'Only... obs on iNat - need 60 for CV update'

Posted by dianastuder 11 months ago

This always makes my day! It’s amazing the updates are so regular!

Posted by yayemaster 11 months ago

Nice to see Arthonia-atra made it to the model and the model is now released almost monthly.

Posted by rudolphous 11 months ago

And motivating nudges on obs where the taxon is closing in on CV
'Only... obs on iNat - need 60 for CV update'

@dianastuder: Glad to see I'm not the only identifier encouraging observers to bring species over the CV threshold!

Posted by rupertclayton 11 months ago

It's always very satisfying to see species that I've observed make it up to the Computer Vision level. It encourages me to take photos of lots of individuals, not just one or two.

Posted by sullivanribbit 11 months ago

CV AI is nice but I hope that the AI never gets as good as human botanists to correctly identify plant species. We still need to be able to have viable careers in botany. I enjoy learning to identify plant species from the photos, but just hope that I am not ultimately training a computer to take my job some day. Something to consider . . .

Posted by reodell 11 months ago

This is great! Is there a place to download the model, especially in ONNX format? Thanks!

Posted by codrin_bucur 9 months ago

Here's the link to the v2.1 update, which contains loarie's comment where he includes the links to updates from v1.0 to v1.6
After v1.6 is v2.1. :) --- https://www.inaturalist.org/posts/75633-a-new-computer-vision-model-v2-1-including-1-770-new-taxa

Posted by finding_remo 6 months ago

Add a Comment

Sign In or Sign Up to add comments