• Àüü
  • ÀüÀÚ/Àü±â
  • Åë½Å
  • ÄÄÇ»ÅÍ
´Ý±â

»çÀÌÆ®¸Ê

Loading..

Please wait....

±¹³» ³í¹®Áö

Ȩ Ȩ > ¿¬±¸¹®Çå > ±¹³» ³í¹®Áö > Çѱ¹Á¤º¸°úÇÐȸ ³í¹®Áö > Á¤º¸°úÇÐȸ ÄÄÇ»ÆÃÀÇ ½ÇÁ¦ ³í¹®Áö (KIISE Transactions on Computing Practices)

Á¤º¸°úÇÐȸ ÄÄÇ»ÆÃÀÇ ½ÇÁ¦ ³í¹®Áö (KIISE Transactions on Computing Practices)

Current Result Document : 8 / 10 ÀÌÀü°Ç ÀÌÀü°Ç   ´ÙÀ½°Ç ´ÙÀ½°Ç

ÇѱÛÁ¦¸ñ(Korean Title) ¼øȯ½Å°æ¸ÁÀ» ÀÌ¿ëÇÑ ÇÑ±Û Çʱâü ÀνÄ
¿µ¹®Á¦¸ñ(English Title) Hangul Handwriting Recognition using Recurrent Neural Networks
ÀúÀÚ(Author) ±èº´Èñ   À庴Ź   Byoung-Hee Kim   Byoung-Tak Zhang  
¿ø¹®¼ö·Ïó(Citation) VOL 23 NO. 05 PP. 0316 ~ 0321 (2017. 05)
Çѱ۳»¿ë
(Korean Abstract)
¿Â¶óÀÎ ¹æ½ÄÀÇ ÇÑ±Û Çʱâü ÀÎ½Ä ¹®Á¦¸¦ ºÐ¼®ÇÏ°í ¼øȯ½Å°æ¸Á ±â¹ÝÀÇ ÇعýÀ» ¸ð»öÇÑ´Ù. Çѱ۳¹±ÛÀÚ ÀÎ½Ä ¹®Á¦¸¦ ¼ø¼­µ¥ÀÌÅÍ ·¹ÀÌºí¸µÀÇ °üÁ¡¿¡¼­ ¼­¿­ ºÐ·ù, ±¸°£ ºÐ·ù, ½Ã°£º° ºÐ·ùÀÇ ¼¼ ´Ü°è·Î ±¸ºÐÇÏ¿© °¢°¢¿¡ ´ëÇÑ ÇعýÀ» »ìÆ캸¸ç, ÇѱÛÀÇ ±¸¼º ¿ø¸®¸¦ °í·ÁÇÑ ÇØ°á ¹æ¾ÈÀ» Á¤¸®ÇÑ´Ù. ÇÑ±Û 2350±ÛÀÚ¿¡ ´ëÇÑ ¿Â¶óÀÎ Çʱâü µ¥ÀÌÅÍ¿¡ GRU(gated recurrent unit)ÀÇ ´ÙÃþ ±¸Á¶¸¦ °¡Áö´Â ¼­¿­ ºÐ·ù¸ðµ¨À» Àû¿ëÇÑ °á°ú, ³¹±ÛÀÚ ÀÎ½Ä Á¤È®µµ´Â 86.2%, ÃÊ・Áß・Á¾¼º ±¸¼º¿¡ µû¸¥ 6°¡Áö À¯Çü ºÐ·ù Á¤È®µµ´Â 98.2%·Î ÃøÁ¤µÇ¾ú´Ù. À¯Çü ºÐ·ù ¸ðµ¨·Î ȹÀÇ ÁøÇà¿¡ µû¸¥ À¯Çü º¯È­ ¿ª½Ã ³ôÀº Á¤È®µµ·Î ÀνÄÇÏ´Â °á°ú¸¦ ÅëÇØ, ¼øȯ½Å°æ¸ÁÀ» ÀÌ¿ëÇÏ¿© ¼ø¼­ µ¥ÀÌÅÍ¿¡¼­ ÇѱÛÀÇ ±¸Á¶¿Í °°Àº °íÂ÷¿øÀû Áö½ÄÀ» ÇнÀÇÒ ¼ö ÀÖÀ½À» È®ÀÎÇÏ¿´´Ù.
¿µ¹®³»¿ë
(English Abstract)
We analyze the online Hangul handwriting recognition problem (HHR) and present solutions based on recurrent neural networks. The solutions are organized according to the three kinds of sequence labeling problem - sequence classifications, segment classification, and temporal classification, with additional consideration of the structural constitution of Hangul characters. We present a stacked gated recurrent unit (GRU) based model as the natural HHR solution in the sequence classification level. The proposed model shows 86.2% accuracy for recognizing 2350 Hangul characters and 98.2% accuracy for recognizing the six types of Hangul characters. We show that the type recognizing model successfully follows the type change as strokes are sequentially written. These results show the potential for RNN models to learn high-level structural information from sequential data.
Å°¿öµå(Keyword) ¼øȯ½Å°æ¸Á   ÇѱÛÇʱâü ÀνĠ  µö·¯´× ÀÀ¿ë   ÇÑ±Û ±¸¼º ¿ø¸®¸¦ °í·ÁÇÑ ÀνĠ  ¿Â¶óÀÎ Çʱâü ÀνĠ  recurrent neural networks   Hangul handwriting recognition   deep learning application   online handwriting recognition  
ÆÄÀÏ÷ºÎ PDF ´Ù¿î·Îµå