Using a Statistical Language Model to Improve the Performance of an HMM-Based Cursive Handwriting Recognition System

By Urs-Viktor Marti, Horst Bunke. International journal of pattern recognition and artificial intelligence, 2001

491 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Media Technology

View Paper: https://www.semanticscholar.org/paper/e15725948c2ea8b190b825a0887e430dc4898428

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Media Technology

Citations in the corpus

Bidirectional Language Model for Handwriting Recognition (2012, unknown): “For a more detailed description of the normalization and feature extraction, we refer to [10].”
Long-short term memory neural networks language modeling for handwriting recognition (2012, methodology): “It has been shown that it is beneficial to not only take the input image into account when estimating the most likely word sequence, but also the target language [9].”
Improving Offline Handwritten Text Recognition with Hybrid HMM/ANN Models (2011, unknown): “Handwritten image normalization from a scanned image includes several steps, which usually begin with image cleaning, page skew correction, and line detection [9].”
The IAM-database: an English sentence database for offline handwriting recognition (2002, methodology): “The database described in this paper was instrumental in our research on unconstrained English sentence recognition using statistical language models [12,16,17,13,14].”
Dropout Improves Recurrent Neural Networks for Handwriting Recognition (2013, methodology): “The HMM is also represented as an FST H and is composed with the lexicon FST L , and the language model G .”
IAM-OnDB - an on-line English sentence database acquired from handwritten text on a whiteboard (2005, methodology): “First, the existing recognizer [14] has been designed for of line data, and secondly, it is straightforward to convert on line data to the off-line modality.”
Text-Independent Writer Identification and Verification Using Textural and Allographic Features (2007, methodology): “Schlapbach and Bunke [31] use HMM-based handwriting recognizers [32] for writer identification and verification .”
Offline recognition of unconstrained handwritten texts using HMMs and statistical language models (2004, background): “To our knowledge, the recognition of unconstrained texts has been addressed in very fewworks [9], [10].”
Supervised Sequence Labelling with Recurrent Neural Networks (2012, methodology): “Substantial preprocessing was used for this task (Marti and Bunke, 2001).”
Word spotting for historical documents (2007, background): “In [18], the authors discuss the application of a Hidden Markov model for recognizing handwritten material that was produced specifically for this purpose.”