2017 14th IAPR International Conference on Document Analysis and Recognition (ICDAR) 2017
DOI: 10.1109/icdar.2017.110
|View full text |Cite
|
Sign up to set email alerts
|

Data Augmentation for Recognition of Handwritten Words and Lines Using a CNN-LSTM Network

Abstract: Agradezco a mis chicos que siempre me tuvieron paciencia cuando no rendía como esperaba, agradezco a mis padres que nunca me presionaron por llevar el ritmo con el que me sentía cómodo, agradezco a la universidad por permitirme convertirme en profesional dentro de ella y agradezco a cada persona que de alguna manera llenó de obstáculos mi proceso académico porque en su superación encontré mi realización. Agradezco a mi director de tesis por su paciencia y agradezco a internet, porque en él conecté con una frac… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1
1

Citation Types

0
103
0
1

Year Published

2018
2018
2021
2021

Publication Types

Select...
5
3
2

Relationship

0
10

Authors

Journals

citations
Cited by 154 publications
(104 citation statements)
references
References 17 publications
0
103
0
1
Order By: Relevance
“…In the paper [64], two data expansion and normalization techniques were presented, namely, a novel profile normalization technique for both word and line images and an extension of existing text images using random perturbations on a regular grid. These techniques, combined with LSTM CNN, significantly reduce error rates when recognizing the handwriting in characters and words.…”
Section: Literature Reviewmentioning
confidence: 99%
“…In the paper [64], two data expansion and normalization techniques were presented, namely, a novel profile normalization technique for both word and line images and an extension of existing text images using random perturbations on a regular grid. These techniques, combined with LSTM CNN, significantly reduce error rates when recognizing the handwriting in characters and words.…”
Section: Literature Reviewmentioning
confidence: 99%
“…Neural Networks (CNNs) [14] as a combination of convolutional and fully connected neural networks applied with data augmentation and training techniques [15]. The AlexNet was the winner of the ImageNet Large Scale Visual Recognition Challenge 2012 [16], and continues to be the source of inspiration for winners in years thereafter.…”
Section: Imagenet Classification With Deep Convolutionalmentioning
confidence: 99%
“…To artificially increase the amount of training data, we augment the existing preprocessed images by applying minor alterations to them. To simulate naturally occuring variations in handwritten text line images, we combine dilation, erosion and grid-like distortions [37]. These methods are applied to the original line image randomly with an independent probability of 50 %.…”
Section: E Preprocessing and Data Augmentationmentioning
confidence: 99%