DVHMM: variable length text recognition error model

DVHMM: variable length text recognition error model
复制标题

DVHMM:变长文本识别错误模型

DOI:
10.1109/icpr.2002.1047807
复制
发表时间:
2002
期刊:
Object recognition supported by user interaction for service robots
影响因子:
--
通讯作者:
K. Aihara
K. Aihara
中科院分区:
--
文献类型:
--
作者:
A. Takasu;K. Aihara

文献摘要

被引文献

相似文献

本文提出了一种文本识别误差模型--双变长输出隐马尔可夫模型(DVHMM),并给出了一种基于EM算法的参数估计算法。尽管现有的概率错误模型限于替换(1,1)、插入(1,0)和删除(0,1)错误,但是DVHMM可以处理包括替换、插入和删除的任何长度对(i,j)的错误模式。
This paper proposes a text recognition error model called the dual variable length output hidden Markov model (DVHMM) and gives a parameter estimation algorithm based on the EM algorithm. Although existing probabilistic error models are limited to substitution (1, 1), insertion (1, 0), and deletion (0, 1) errors, the DVHMM can handle error patterns of any pair (i, j) of lengths including substitution, insertion, and deletion.