It might be cool to provide a badge of some kind on an observation page if one of the observation photos was incorporated in one of the iNat contest datasets. This could help users feel good about their iNat contributions helping science, and it could provide a way for us and for the ML researchers to thank and credit the individuals who've helped improve iNat, the visipedia team, and the worldwide computer vision community.
Not sure if we still have the linkage between the photos as they exist on iNat and the obfuscated photo names that were provided in the datasets. If not, we could use a combo of scripting, data munging, and manual digging to do some manual data archaeology using the data that is in the contest datasets: username, taxon, photo & its features, date time, etc to try to track down the original observation for each photo in the dataset.
Once we have a list of photo ids that were in the contest dataset, here are a few ways we could leverage existing iNat features to give users credit or a connection to the contests without building too much new code:
We could make a project for a CV contest, add the photos to that project, and then users who join the project would see the CV contest project name on their observation page. One benefit of this approach is that projects have journals which Grant & co could use to communicate with the observers who've joined the project.
Another way to do this would be to use the data partners system in iNat to make the CV contests a data partner. One benefit to this approach is that users wouldn't need to join the project to see the contest name as a badge on their observation page. Another benefit is that it would automatically or easily roll up into the Year in Review.
Another way we could do this might be observation fields.
It might be cool to provide a badge of some kind on an observation page if one of the observation photos was incorporated in one of the iNat contest datasets. This could help users feel good about their iNat contributions helping science, and it could provide a way for us and for the ML researchers to thank and credit the individuals who've helped improve iNat, the visipedia team, and the worldwide computer vision community.
Not sure if we still have the linkage between the photos as they exist on iNat and the obfuscated photo names that were provided in the datasets. If not, we could use a combo of scripting, data munging, and manual digging to do some manual data archaeology using the data that is in the contest datasets: username, taxon, photo & its features, date time, etc to try to track down the original observation for each photo in the dataset.
Once we have a list of photo ids that were in the contest dataset, here are a few ways we could leverage existing iNat features to give users credit or a connection to the contests without building too much new code: