Paired Image to Image Translation for Strikethrough Removal From Handwritten Words
This addresses the challenge of transcribing struck-through handwritten words for applications like genetic criticism, but it is incremental as it builds on existing paired translation methods with architectural improvements.
The paper tackled the problem of removing strikethrough strokes from handwritten words using paired image-to-image translation, and the result showed that the proposed models outperformed the CycleGAN-based state-of-the-art while using less than a sixth of the trainable parameters.
Transcribing struck-through, handwritten words, for example for the purpose of genetic criticism, can pose a challenge to both humans and machines, due to the obstructive properties of the superimposed strokes. This paper investigates the use of paired image to image translation approaches to remove strikethrough strokes from handwritten words. Four different neural network architectures are examined, ranging from a few simple convolutional layers to deeper ones, employing Dense blocks. Experimental results, obtained from one synthetic and one genuine paired strikethrough dataset, confirm that the proposed paired models outperform the CycleGAN-based state of the art, while using less than a sixth of the trainable parameters.