NAF (National Archives Forms Dataset)
Introduced by Davis et al. in Deep Visual Template-Free Form Parsing
This dataset was created with images provided by the United States National Archive and FamilySearch.
The goal of this data is to capture relationships between text/handwriting entities on form images. It will include transcriptions in the future, but doesn't currently.
The form images are organized into "groups", each group containing images of the same form type.
Source: NAF DatasetPapers
| Paper | Code | Results | Date | Stars |
|---|
