A MDRNN-SVM hybrid model for cursive offline handwriting recognition. Bezerra, B., Zanchettin, C., & De Andrade, V. Volume 7553 LNCS , 2012.
doi  abstract   bibtex   
This paper presents a recurrent neural networks applied to handwriting character recognition. The method Multi-dimensional Recurrent Neural Network is evaluated against classical techniques. To improve the model performance we propose the use of specialized Support Vector Machine combined whit the original Multi-dimensional Recurrent Neural Network in cases of confusion letters. The experiments were performed in the C-Cube database and compared with different classifiers. The hierarchical combination presented promising results. © 2012 Springer-Verlag.
@book{
 title = {A MDRNN-SVM hybrid model for cursive offline handwriting recognition},
 type = {book},
 year = {2012},
 source = {Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)},
 volume = {7553 LNCS},
 issue = {PART 2},
 id = {6b68800f-c7a4-3d31-9a6c-7432acc50567},
 created = {2019-02-14T18:02:01.291Z},
 file_attached = {false},
 profile_id = {74e7d4ea-3dac-3118-aab9-511a5b337e8f},
 last_modified = {2019-02-14T18:02:01.291Z},
 read = {false},
 starred = {false},
 authored = {true},
 confirmed = {false},
 hidden = {false},
 private_publication = {false},
 abstract = {This paper presents a recurrent neural networks applied to handwriting character recognition. The method Multi-dimensional Recurrent Neural Network is evaluated against classical techniques. To improve the model performance we propose the use of specialized Support Vector Machine combined whit the original Multi-dimensional Recurrent Neural Network in cases of confusion letters. The experiments were performed in the C-Cube database and compared with different classifiers. The hierarchical combination presented promising results. © 2012 Springer-Verlag.},
 bibtype = {book},
 author = {Bezerra, B.L.D. and Zanchettin, C. and De Andrade, V.B.},
 doi = {10.1007/978-3-642-33266-1_31}
}

Downloads: 0