There’s this topic:
Help clean up computer vision errors!
Add species to this list that are frequently getting misidentified based on computer vision (CV) suggestions. This is a “clean-up” list. Only include species that the community is having difficulty keeping up with , such as those with few people making identifications or generally overlooked groups. As a rough rule of thumb, the species should have at least ~100 observations that are currently misidentified. This topic isn’t for species that are already being well handled by the community reidentifying them. That said, if you are feeling overwhelmed, appear to be alone in the effort trying to reidentify, or just dang tired of fixing the IDs, I think it might qualify for inclusion here.
And this is already on the staff’s to-do list:
Continue to improve the Computer Vision model and the use of geospatial data in suggestions.
Unfortunately the staff have already already said they aren’t planning on moving forward with a flag or alert for fixing computer vision errors like this; see this closed feature request: Add a flag for frequent computer vision identification errors
5 Likes