TorMentor: Data Augmentation for Historical Documents

Poster at CVPR 2022

We propose the use of fractals as mean of efficient data augmentation. Specifically, we employ plasma fractals as a means of adapting global image augmentation transformations into continuous local transforms. We formulate the diamond square algorithm as a cascade of simple convolution operations allowing efficient computation of plasma fractals on the GPU. We present the TorMentor image augmentation framework that is totally modular and deterministic across images and point-clouds. All image augmentation operations can be combined through pipelining and random branching to form flow networks of arbitrary width and depth. We demonstrate the efficiency of the proposed approach with experiments on document image segmentation (binarization) with the DIBCO datasets. The proposed approach demonstrates superior performance to traditional image augmentation techniques. Finally, we use extended synthetic binary text images in a self-supervision regiment and outperform the same model when trained with limited data and simple extensions.

Anguelos Nicolaou, Vincent Christlein, Edgar Riba, Jian Shi, Georg Vogeler, Mathias Seuret; pushlied in the Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, 2022, pp. 2707-2711

CVPR 2022 Repository entry



Cite this blog post
anguelos (2022, July 2). TorMentor: Data Augmentation for Historical Documents. DiDip. Retrieved April 24, 2024, from https://didip.hypotheses.org/2193

anguelos

I am a Computer Vision Postdoc Researcher.

You may also like...

Search OpenEdition Search

You will be redirected to OpenEdition Search