Minimum Risk Training for Handwritten Chinese/Japanese Text Recognition Using Semi-Markov Conditional Random Fields

ISCAS OpenIR

	Minimum Risk Training for Handwritten Chinese/Japanese Text Recognition Using Semi-Markov Conditional Random Fields
	Zhou, Xiang-Dong; Tian, Feng; Liu, Cheng-Lin; Wang, Hong-An
	2013
会议名称	12th International Conference on Document Analysis and Recognition (ICDAR)
页码	940-944
会议日期	AUG 25-28, 2013
会议地点	Washington, DC
收录类别	CPCI
出版地	IEEE
ISSN	1520-5363
部门归属	[Zhou, Xiang-Dong; Tian, Feng; Wang, Hong-An] Chinese Acad Sci, Beijing Key Lab Human Comp Interact, Inst Software, Beijing 100190, Peoples R China.
摘要	Semi-Markov conditional random fields (semi-CRFs) are usually trained with maximum a posteriori (MAP) criterion which adopts the 0/1 cost for measuring the loss of misclassification. In this paper, based on our previous work on handwritten Chinese/Japanese text recognition (HCTR) using semi-CRFs, we propose an alternative parameter learning method by minimizing the risk, in which the misclassification costs are not equal, but different depending on the hypothesis and the ground-truth. The proposed method is lattice-based, i.e., the hypothesis space is the entire lattice on which the semi-CRF is defined. Experimental results on two online handwriting databases: CASIA-OLHWDB and TUAT Kondate demonstrate that minimum-risk training can yield superior string recognition rates compared to MAP training.; Semi-Markov conditional random fields (semi-CRFs) are usually trained with maximum a posteriori (MAP) criterion which adopts the 0/1 cost for measuring the loss of misclassification. In this paper, based on our previous work on handwritten Chinese/Japanese text recognition (HCTR) using semi-CRFs, we propose an alternative parameter learning method by minimizing the risk, in which the misclassification costs are not equal, but different depending on the hypothesis and the ground-truth. The proposed method is lattice-based, i.e., the hypothesis space is the entire lattice on which the semi-CRF is defined. Experimental results on two online handwriting databases: CASIA-OLHWDB and TUAT Kondate demonstrate that minimum-risk training can yield superior string recognition rates compared to MAP training.
语种	英语
内容类型	会议论文
URI标识	http://ir.iscas.ac.cn/handle/311060/16509
专题	中国科学院软件研究所
推荐引用方式 GB/T 7714	Zhou, Xiang-Dong,Tian, Feng,Liu, Cheng-Lin,et al. Minimum Risk Training for Handwritten Chinese/Japanese Text Recognition Using Semi-Markov Conditional Random Fields[C]. IEEE,2013:940-944.