@inproceedings{493d0915602148138a128fc375e39e54,
title = "Maximum-minimum similarity training for text extraction",
abstract = "In this paper, the discriminative training criterion of maximum-minimum similarity (MMS) is used to improve the performance of text extraction based on Gaussian mixture modeling of neighbor characters. A recognizer is optimized in the MMS training through maximizing the similarities between observations and models from the same classes, and minimizing those for different classes. Based on this idea, we define the corresponding objective function for text extraction. Through minimizing the objective function by using the gradient descent method, the optimum parameters of our text extraction method are obtained. Compared with the maximum likelihood estimation (MLE) of parameters, the result trained with the MMS method makes the overall performance of text extraction improved greatly. The precision rate decreased little from 94.59% to 93.56%, but the recall rate increased a lot from 80.39% to 98.55%.",
author = "Hui Fu and Xiabi Liu and Yunde Jia",
year = "2006",
doi = "10.1007/11893295_31",
language = "English",
isbn = "3540464840",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "268--277",
booktitle = "Neural Information Processing - 13th International Conference, ICONIP 2006, Proceedings",
address = "Germany",
note = "13th International Conference on Neural Information Processing, ICONIP 2006 ; Conference date: 03-10-2006 Through 06-10-2006",
}