Segmental K-Means Learning with Mixture Distribution for HMM Based Handwriting Recognition
This paper investigates the performance of hidden Markov models (HMMs) for handwriting recognition. The Segmental K-Means algorithm is used for updating the transition and observation probabilities, instead of the Baum-Welch algorithm. Observation probabilities are modelled as multi-variate Gaussian mixture distributions. A deterministic clustering technique is used to estimate the initial parameters of an HMM. Bayesian information criterion (BIC) is used to select the topology of the model. The wavelet transform is used to extract features from a grey-scale image, and avoids binarization of the image.
KeywordsHide Markov Model Bayesian Information Criterion Mixture Component Mixture Distribution Handwriting Recognition
- 1.Rabiner, L.: A tutorial on hidden markov models and selected applications in speech recognition. Trans. on. IEEE 77(2), 257–286 (1989)Google Scholar
- 6.Su, T., Dy, J.: A deterministic method for initializing k-means clustering. In: Proc. of The 16th IEEE Int. Conf. on Tools with Artificial Intelligence, pp. 784–786 (2004)Google Scholar
- 7.Rahimi, A.: An erratum for a tutorial on hidden markov models and selected applications in speech recognition. In: Online article (2000)Google Scholar
- 9.Biem, A.: A model selection criterion for classification: Application to hmm topology optimization. In: Proceedings of the 7th ICDAR, pp. 104–108 (2003)Google Scholar