Abstract
We present a writer-independent system for online handwriting recognition that can handle a variety of writing styles including cursive script and handprinting. The input to our system contains the pen trajectory information, encoded as a time-ordered sequence of feature vectors. A time-delay neural network is used to estimate a posteriori probabilities for characters in a word. A hidden Markov model segments the word in a way that optimizes the global word score, using a dictionary in the process. A geometrical normalization scheme and a fast but efficient dictionary search are also presented. Trained on 20 k words from 59 writers, using a 25 k-word dictionary, our system reached recognition rates of 89% for characters and 80% for words on test data from a disjoint set of writers.
Similar content being viewed by others
References
Bengio Y, LeCun Y, Henderson D (1994) Globally trained handwritten word recognizer using spatial representation, convolutional neural networks and hidden markov models. In: Cowan JD, Tesauro G, Alspector J (eds) Advances in neural information processing systems 6. Morgan Kauffman, Denver, Colo., pp 937–944
Berthod M (1990) On-line analysis of cursive writing. In: Suen CY, Mori RD (eds) Computer analysis and perception: Vol. 1 — Visual Signals, CRC Press, Boca Raton, Florida, pp 55–81
Bourlard H, Wellekens C (1990) Links between markov models and multilayer perception. IEEE Trans Patt Anal Machine Intell 12:1167–1178
Bourlard H, Morgan N (1993) Connectionist speech recognition. Kluver Academic, Boston, MA
Enrich RW, Koehler KJ (1975) Experiments in the contextual recognition of cursive script. IEEE Trans Comput 24:182–194
Fujisaka T, Nathan K, Cho W, Beigi H (1993) On-line unconstrained handwriting recognition by a probabilistic method. International Workshop on Frontiers in Handwriting Recognition III. Buffalo, N.Y., pp 231–241
Gupta VN, Lennig M, Mermelstein P (1988) Fast search strategy in a large vocabulary word recognizer. J Acoust Soc Am 84:2007–2017
Guyon I, Albrecht P, Cun YL, Denker J, Hubbard W (1991) Design of a neural network character recognizer for a touch terminal. Patt Recogn 24:105–119
Ha JY, Oh SC, Kim JH, Kwon YB (1993) Unconstrained handwritten word recognition with interconnected hidden markov models. International Workshop on Frontiers in Handwriting Recognition III, Buffalo, N.Y., pp 455–460
Harmon LD (1962) Automatic reading of cursive script. In: Fisher GL (ed) Optical character recognition, Washington, DC, pp 151–152(A)
Huang WY, Lippmann RP (1990) HMM speech recognition with neural net discrimination. In: Touretzky DS (ed) Advances in neural information processing systems 2, Morgan Kaufmann, Denver, Colo., pp 194–202
Keeler J, Rumelhart DE, Leow WK (1991) Integrated segmentation and recognition of hand-printed numerals In: Lippmann RP, Moody JE, Touretzky DS (eds) Advances in neural information processing systems 3, Morgan Kaufmann, Denver, Colo., pp 557–563
Lang KJ, Hinton G (1988) A time delay neural network architecture for speech recognition. Technical Report CMU-cs-88-152, Carnegie-Mellon University, Pittsburgh, Pa
Lecolinet E, Baret O (1993) Cursive word recognition: methods and strategies. NATO-ASI summer school, fundamentals in Handwriting Recognition, Bonas, France, 1993 July, preprint
LeCun Y (1989) Generalization and network design strategies. In: Pfeifer R, Schreter Z, Fogelman F, Steels L (eds) Connectionism in perspective, Elsevier Zurich
Levenshtein VI (1966) Binary codes capable of correcting deletions, insertions and reversals. Soviet Physics-Doklady, 10:707–710
Matan O, Burges CJC, LeCun Y, Denker J (1992) Multidigit recognition using a space dispacement neural network. In: Moody JE, Hanson SJ, Lippmann RP (eds) Advances in neural information processing systems 4, Morgan Kaufmann, Denver, Colo., pp 488–495
Mermelstein P, Eden M (1964) Experiments on computer recognition of connected handwritten words. Information and Control 7:255
Morasso P, Barberis L, Pagliano S, Vergano D (1993) Recognition experiments of cursive dynamic handwriting with self-organizing networks. Patt Recogn 26:451–460
Nag R, Wong KH, Fallside F (1986) Script recognition using hidden Markov models. IEEE International Conference on Acoustics, Speech, and Signal Processing, Tokyo, pp 2071–2074
Okuda T, Tanaka E, Tamotsu K (1976) A method for the correction of garbled words based on the Levenshtein metric. IEEE Trans Comput 25:172–177
Rabiner LR (1989) A tutorial on hidden Markov models and selected applications in speech recognition. Proc IEEE:77:257–285
Rumelhart DE, Hinton GE, Williams RJ (1986) Learning internal representations by error propagation. In: Parallel distributed processing: explorations in the microstructure of cognition, volume I, Bradford Books, Cambridge, Mass., pp 318–362
Sayre KM (1973) Machine recognition of handwritten words: a project report. Patt Recogn 5:213–228
Schenkel M, Weissman H, Guyon I, Nohl C, Henderson D (1993) Recognition-based segmentation of on-line hand-printed words. In: Hanson SJ, Cowan JD, Giles CL (eds) Advances in neural information processing systems 5, Morgan Kaufmann, Denver, Colo., pp 723–730
Schomaker L (1993) Using stroke- or character-based self-organizing maps in the recognition of on-line, connected cursive script. Patt Recogn 26:443–450
Tappert CC, Suen CY, Wakahara T (1990) The state of the art in on-line handwriting recognition. IEEE Trans Patt Anal Machine Intell 12:787–808
Waibel A, Hanazawa T, Hinton G, Shikano K, Lang K (1989) Phoneme recognition using time-delay neural networks. IEEE Trans Acoustics Speech Signal Processing 37:328–339
Weissman H, Schenkel M, Guyon I, Nohl C, Henderson D (1994) Recognition-based segmentation of on-line handprinted words: input vs. output segmentation. Patt Recogn 27:405–420
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Schenkel, M., Guyon, I. & Henderson, D. On-line cursive script recognition using time-delay neural networks and hidden Markov models. Machine Vis. Apps. 8, 215–223 (1995). https://doi.org/10.1007/BF01219589
Issue Date:
DOI: https://doi.org/10.1007/BF01219589