Topic
Intelligent word recognition
About: Intelligent word recognition is a research topic. Over the lifetime, 2480 publications have been published within this topic receiving 45813 citations.
Papers published on a yearly basis
Papers
More filters
20 Aug 1996
TL;DR: This paper describes the structural learning of Kanji patterns in on-line handwritten character recognition by investigating which subpattern or the pattern as a whole is non-standard, registers the (sub)pattern and extends the effect of the registration to all the character categories whose shapes include it.
Abstract: This paper describes the structural learning of Kanji patterns in on-line handwritten character recognition. Upon the request to learn an input pattern, the system investigates which subpattern or the pattern as a whole is non-standard, registers the (sub)pattern and extends the effect of the registration to all the character categories whose shapes include it. A character pattern representation dictionary stores character patterns as combinations of subpatterns so that a common subpattern is shared by all the character categories which include it in their shapes. The recognizer constructs all template patterns from its constituent subpatterns for matching them with an input pattern. Registration of a character pattern invokes identification and registration of a non-standard subpattern in it so that the effect extends to all the characters whose shapes include it. A preliminary evaluation shows it is highly effective without any bad side effect.
14 citations
23 Aug 2004
TL;DR: By combining the character-level trained classifier and the string-level training, this work has achieved higher string recognition performance and shows the experimental results of three classifier structures on the numeral strings of NIST Special Database 19.
Abstract: The performance of handwritten numeral string recognition integrating segmentation and classification relies on the classification accuracy and the resistance to non-characters of the underlying classifier. The classifier can be trained at either character level (with character and non-character samples) or string level (with string samples). We show that both character-level and string-level training yield superior string recognition performance. String-level training improves segmentation but deteriorates classification. By combining the character-level trained classifier and the string-level trained classifier, we have achieved higher string recognition performance. We show the experimental results of three classifier structures on the numeral strings of NIST Special Database 19.
14 citations
01 Jan 2002
TL;DR: The evaluation results for lexicon-free handwriting recognition demonstrate the effectiveness of the proposed methods.
Abstract: A system for off-line handwritten text recognition is presented. It is characterized by a segmentation-free approach, i.e. whole lines of text are processed by the recognition module. The methods used for pre-processing, feature extraction, and statistical modelling are described, and several experiments on writer-independent, multiple writer, and single writer handwriting recognition tasks are conducted. Particularly, the incorporation of linear discriminant analysis, allograph character models, and statistical language knowledge is investigated. The evaluation results for lexicon-free handwriting recognition demonstrate the effectiveness of the proposed methods.
14 citations
01 Jan 1998
TL;DR: It is demonstrated that training a simple recurrent network to activate a representation of all the words in a sequence allows the network to learn to recognise onset-embedded words without requiring a training set that is already lexically segmented.
Abstract: Onset-embedded words (e.g. cap in captain) present a problem for accounts of spoken word recognition since information coming after the offset of the embedded word may be required for identification. We demonstrate that training a simple recurrent network to activate a representation of all the words in a sequence allows the network to learn to recognise onset-embedded words without requiring a training set that is already lexically segmented. We discuss the relationship between our model and other accounts of lexical segmentation and word recognition, and compare the model’s performance to psycholinguistic data on the recognition of onset-embedded words.
13 citations
01 Sep 2015
TL;DR: The work described in this paper presents efficiency of Zernike moments over Hu's seven moment with zoning for automatic recognition of handwritten `MODI' characters.
Abstract: HOCR is abbreviated as Handwritten Optical Character Recognition. HOCR recognizes handwritten characters from a digital image of documents. Shape identification and feature extraction is very important part of any OCR. Feature extraction defines shape of the character as precisely and as uniquely as possible. Zernike moments describes shape, identify rotation invariant due to its orthogonal property. ‘MODI’ is an ancient script of India had cursive and complex representation of characters. The work described in this paper presents efficiency of Zernike moments over Hu's seven moment with zoning for automatic recognition of handwritten ‘MODI’ characters. 82.61% recognition rate was achieved by using zone based approach for Zernike moments.
13 citations