An Easy-To-Use Recurrent Neural Network Architecture for Sequence Recognition

Marcus Liwicki

In: Technical Committee on Pattern Recognition and Media Understanding. Pattern Recognition and Media Understanding (PRMU-2010) February 18-19 Tokyo Japan Seiten 69-74 2/2010.


In this presentation the recently introduced Bidirectional Recurrent Neural Networks will be described. They allow for a direct recognition of raw pixel data. Promising experimental results on various datasets from different countries are presented. A toolkit implementing the networks is freely available for public.


Deutsches Forschungszentrum für Künstliche Intelligenz
German Research Center for Artificial Intelligence