Search Results for author: Phillip Benjamin Ströbel

Found 2 papers, 2 papers with code

Transformer-based HTR for Historical Documents

1 code implementation • 21 Mar 2022 • Phillip Benjamin Ströbel, Simon Clematide, Martin Volk, Tobias Hodel

We apply the TrOCR framework to real-world, historical manuscripts and show that TrOCR per se is a strong model, ideal for transfer learning.

HTR Transfer Learning

Paper
Code

Evaluation of HTR models without Ground Truth Material

1 code implementation • LREC 2022 • Phillip Benjamin Ströbel, Simon Clematide, Martin Volk, Raphael Schwitter, Tobias Hodel, David Schoch

The evaluation of Handwritten Text Recognition (HTR) models during their development is straightforward: because HTR is a supervised problem, the usual data split into training, validation, and test data sets allows the evaluation of models in terms of accuracy or error rates.

Handwritten Text Recognition HTR +1

Paper
Code

Cannot find the paper you are looking for? You can Submit a new open access paper.