POPP Generic dataset: 1926 Paris census.
Constum, Kempf, Paquet, Tranouez, Chatelain, Brée, and Merveille (2022).
Recognition and Information Extraction in Historical Handwritten Tables:
Toward Understanding Early 20th Century Paris Census. DAS, pp. 143–157.
Source: https://doi.org/10.5281/zenodo.6581158
License: CC BY 4.0, https://creativecommons.org/licenses/by/4.0/
Changes: six pages cropped to columns 6–15 and stored as lossless PNG;
180 source transcriptions parsed into ten fields with whitespace collapsed.
Original annotations and source coordinates are retained in labels.csv.
Model responses are predictions, not source transcriptions.