Separating Optical and Language Models through Encoder-Decoder Strategy for Transferable Handwriting Recognition - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2018

Separating Optical and Language Models through Encoder-Decoder Strategy for Transferable Handwriting Recognition

Résumé

Lack of data can be an issue when beginning a new study on historical handwritten documents. To deal with this, we propose a deep-learning based recognizer which separates the optical and the language models in order to train them separately using different resources. In this work, we present the optical encoder part of a multilingual transductive transfer learning applied to historical handwriting recognition. The optical encoder transforms the input word image into a non-latent space that depends only on the letter-n-grams: it enables it to be independent of the language. This transformation avoids embedding a language model and operating the transfer learning across languages using the same alphabet. The language decoder creates from a vector of letter-n-grams a word as a sequence of characters. Experiments show that separating optical and language model can be a solution for multilingual transfer learning.
Fichier principal
Vignette du fichier
icfhr(2).pdf (239.17 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-01821598 , version 1 (22-06-2018)

Identifiants

  • HAL Id : hal-01821598 , version 1

Citer

Adeline Granet, Emmanuel Morin, Solen Quiniou, Christian Viard-Gaudin, Harold Mouchère. Separating Optical and Language Models through Encoder-Decoder Strategy for Transferable Handwriting Recognition. 16th International Conference on Frontiers in Handwriting Recognition (ICFHR), Aug 2018, Niagara Falls, Canada. ⟨hal-01821598⟩
198 Consultations
131 Téléchargements

Partager

Gmail Facebook X LinkedIn More