Great work with this project. I was hoping you could either share or explain your workflow for the Rekognition face comparisons, because the current list of people appears to be incomplete. For example, documents 3400, 3478, 3479, 3480, 3481, 3482, 3510, 3824, 3825 (and likely many more) all appear to show the same person (Jean-Luc Brunel).
If the difficult part is identifying detected people, maybe having some kind of view that shows the top unidentified people who appear in the most images, so researchers can help name them.
Great work with this project. I was hoping you could either share or explain your workflow for the Rekognition face comparisons, because the current list of people appears to be incomplete. For example, documents 3400, 3478, 3479, 3480, 3481, 3482, 3510, 3824, 3825 (and likely many more) all appear to show the same person (Jean-Luc Brunel).
If the difficult part is identifying detected people, maybe having some kind of view that shows the top unidentified people who appear in the most images, so researchers can help name them.