A CNN handwritten character recognizer
この論文をさがす
説明
<jats:title>Abstract</jats:title><jats:p>CNNs are used for feature detection in handwritten character recognition. Detected features are fed to a simple classifier network. Performance was tested by using two well‐known ETL data base series: (i) ETL3 consisting of numerals, alphabets and several symbols and (ii) ETL8B2 consisting of Japanese Hirakana characters. the average recognition rate for ETL3 is 94.8%, while that for ETL8B2 is 85.7%. Both series include ‘hard’ characters so distorted that even humans cannot recognize them.</jats:p>
収録刊行物
-
- International Journal of Circuit Theory and Applications
-
International Journal of Circuit Theory and Applications 20 601-612, 1992-09-01
Wiley