Transfer Learning for a Letter-Ngrams to Word Decoder in the Context of Historical Handwriting Recognition with Scarce Resources.
Adeline GranetEmmanuel MorinHarold MouchèreSolen QuiniouChristian Viard-GaudinPublished in: COLING (2018)
Keyphrases
- transfer learning
- handwriting recognition
- scarce resources
- word recognition
- historical documents
- historical manuscripts
- word segmentation
- n gram
- knowledge transfer
- handwritten document images
- word spotting
- handwritten words
- labeled data
- machine learning
- cross domain
- speech recognition
- text classification
- learning algorithm
- reinforcement learning
- resource allocation
- handwritten documents
- transfer knowledge
- semi supervised learning
- context sensitive
- character recognition
- machine learning algorithms
- unlabeled data
- cross lingual
- active learning
- pattern recognition
- language independent
- target domain
- text mining
- resource availability
- association rules