Recognition of handwritten Latin characters with diacritics using CNN
Convolutional Neural Networks (CNN) have achieved huge popularity in solving problems in image analysis and in text recognition. In this work, we assess the effectiveness of CNN-based architectures where a network is trained in recognizing handwritten characters based on Latin script. European langu...
Main Authors: | Edyta Lukasik, Malgorzata Charytanowicz, Marek Milosz, Michail Tokovarov, Monika Kaczorowska, Dariusz Czerwinski, Tomasz Zientarski |
---|---|
Format: | Article |
Language: | English |
Published: |
Polish Academy of Sciences
2021-01-01
|
Series: | Bulletin of the Polish Academy of Sciences: Technical Sciences |
Subjects: | |
Online Access: | https://journals.pan.pl/Content/119099/PDF/12_01819_Bpast.No.69(1)_13.01.21_K1_A.pdf |
Similar Items
-
EFFICIENCY COMPARISON OF NETWORKS IN HANDWRITTEN LATIN CHARACTERS RECOGNITION WITH DIACRITICS
by: Edyta ŁUKASIK, et al.
Published: (2023-12-01) -
Diacritics restoration based on word n-grams for Slovak texts
by: Toth Štefan, et al.
Published: (2021-01-01) -
Quranic diacritic and character segmentation and recognition using flood fill and k-nearest neighbors algorithm
by: Alotaibi, Faiz E A L
Published: (2019) -
Morphological, syntactic and diacritics rules for automatic diacritization of Arabic sentences
by: Amine Chennoufi, et al.
Published: (2017-04-01) -
DEVELOPMENT OF AN IMPROVED DATABASE FOR YORUBA HANDWRITTEN CHARACTER
by: OLUWASHINA O. OYENIRAN, et al.
Published: (2021-12-01)