Abstract | ||
---|---|---|
One of the limitations for using Deep Learning models to solve binarization tasks is that there is a lack of large quantities of labeled data available to train such models. Efforts to create synthetic data for binarization mostly rely on heuristic image processing techniques and generally lack realism. In this work, we propose a method to produce realistic synthetic data using an adversarially trained image translation model. We extend the popular CycleGAN model to be conditioned on the ground truth binarization mask as it translates images from the domain of synthetic images to the domain of real images. For evaluation, we train deep networks on synthetic datasets produced in different ways and measure their performance on the DIBCO datasets. Compared to not pretraining, we reduce error by 13% on average, and compared to pretraining on unrealistic data, we reduce error by 6%. Visually, we show that DGT-CycleGAN model produces more realistic synthetic data than other models. |
Year | DOI | Venue |
---|---|---|
2019 | 10.1109/ICDAR.2019.00036 | 2019 International Conference on Document Analysis and Recognition (ICDAR) |
Keywords | Field | DocType |
Binarization,Deep Learning,Generative Adversarial Networks,Synthetic Data | Image translation,Heuristic,Pattern recognition,Computer science,Image processing,Synthetic data,Ground truth,Artificial intelligence,Generative grammar,Deep learning,Real image | Conference |
ISSN | ISBN | Citations |
1520-5363 | 978-1-7281-3015-6 | 0 |
PageRank | References | Authors |
0.34 | 18 | 4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Chris Tensmeyer | 1 | 20 | 4.83 |
Mike Brodie | 2 | 0 | 0.34 |
Daniel Saunders | 3 | 0 | 0.34 |
Tony R. Martinez | 4 | 1364 | 100.44 |