Parsimonious HMMs for Offline Handwritten Chinese Text Recognition

TL;DR

This paper introduces parsimonious HMMs with state tying for offline handwritten Chinese text recognition, reducing model size and decoding time while improving accuracy by leveraging character similarities.

Contribution

The paper proposes a novel state tying approach for HMMs that enhances model efficiency and accuracy in Chinese text recognition, using data-driven algorithms and deep neural networks.

Findings

Achieved 6.2% relative CER reduction with DNN-HMMs.

Reduced model size by 25% and decoding time by 60%.

Outperformed conventional DNN-HMMs in compact settings.

Abstract

Recently, hidden Markov models (HMMs) have achieved promising results for offline handwritten Chinese text recognition. However, due to the large vocabulary of Chinese characters with each modeled by a uniform and fixed number of hidden states, a high demand of memory and computation is required. In this study, to address this issue, we present parsimonious HMMs via the state tying which can fully utilize the similarities among different Chinese characters. Two-step algorithm with the data-driven question-set is adopted to generate the tied-state pool using the likelihood measure. The proposed parsimonious HMMs with both Gaussian mixture models (GMMs) and deep neural networks (DNNs) as the emission distributions not only lead to a compact model but also improve the recognition accuracy via the data sharing for the tied states and the confusion decreasing among state classes. Tested on ICDAR-2013 competition database, in the best configured case, the new parsimonious DNN-HMM can yield a relative character error rate (CER) reduction of 6.2%, 25% reduction of model size and 60% reduction of decoding time over the conventional DNN-HMM. In the compact setting case of average 1-state HMM, our parsimonious DNN-HMM significantly outperforms the conventional DNN-HMM with a relative CER reduction of 35.5%.