Classifying isogenous fields

S. Veeramachaneni1, H. Fujisawa2, Cheng-Lin Liu2, G. Nagy1
1Rensselaer Polytechnic Institute, Troy, NY, USA
2Hitachi Central Research Laboratory Limited, Tokyo, Japan

Tóm tắt

Classifiers that utilize style context in co-occurring patterns increase recognition accuracy. When patterns occur as long isogenous fields, this gain should increase unless negated by parameter estimation errors that increase with field length. We show that our method achieves higher accuracy with longer input fields because it can be trained accurately We also present some ongoing work on simple heuristics to reduce computational complexity of the scheme.

Từ khóa

#Parameter estimation #Pattern recognition #Computational complexity #Laboratories #Electronic mail #Educational institutions #Writing #Instruments #NIST #Random variables

Tài liệu tham khảo

grother, 1995, Handprinted Forms and Character Database NIST Special Database duda, 1973, Pattern Classification and Scene Analysis 10.1109/TIT.1970.1054406 sarkar, 2000, Style Consistency in Pattern Fields 10.1007/s100320200076 10.1109/ICDAR.2001.953774 10.1109/ICDAR.1995.598952 10.1109/ICPR.2002.1048239 10.1109/ICDAR.2001.953969