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

»çÀÌÆ®¸Ê

Loading..

Please wait....

Çмú´ëȸ ÇÁ·Î½Ãµù

Ȩ Ȩ > ¿¬±¸¹®Çå > Çмú´ëȸ ÇÁ·Î½Ãµù > Çѱ¹Á¤º¸Åë½ÅÇÐȸ Çмú´ëȸ > 2018³â Ãß°èÇмú´ëȸ

2018³â Ãß°èÇмú´ëȸ

Current Result Document : 2 / 20 ÀÌÀü°Ç ÀÌÀü°Ç   ´ÙÀ½°Ç ´ÙÀ½°Ç

ÇѱÛÁ¦¸ñ(Korean Title) À¥ À̹ÌÁö ¸¶ÀÌ´×°ú ·£´ý ·¹À̺íÀ» ÀÌ¿ëÇÑ µö·¯´× ±â¹Ý °³ Ç°Á¾ ÀνÄ
¿µ¹®Á¦¸ñ(English Title) Recognition of Dog Breeds based on Deep Learning using a Random-Label and Web Image Mining
ÀúÀÚ(Author) °­¹Î¼®   È«±¤¼®   Min-Seok Kang   Kwang-Seok Hong  
¿ø¹®¼ö·Ïó(Citation) VOL 22 NO. 02 PP. 0201 ~ 0202 (2018. 10)
Çѱ۳»¿ë
(Korean Abstract)
º» ³í¹®¿¡¼­´Â ±âÁ¸ ImageNet[1]°ú Oxford-IIIT Pet Image[2]ÀÇ Dataset¿¡¼­ Á¦°øÇÏ´Â °³ Ç°Á¾ À̹ÌÁö¿Í ÀÎÅÍ³Ý »ó¿¡¼­ °³ Ç°Á¾ À̹ÌÁö¸¦ µ¥ÀÌÅÍ ¸¶ÀÌ´×À» ÅëÇØ È¹µæµÈ °³ Ç°Á¾ À̹ÌÁö¸¦ °áÇÕÇÏ°í Random-LabelÀ» Ãß°¡ ÇÏ¿© °³ Ç°Á¾ 122°³ÀÇ Å¬·¡½º¿Í °³ Ç°Á¾ÀÌ ¾Æ´Ñ 1°³ÀÇ Å¬·¡½º¸¦ ÀνÄÇÏ´Â ¹æ¹ý¿¡ ´ëÇØ ¼Ò°³ ÇÑ´Ù. ±âÁ¸ DB¸¸À» »ç¿ëÇÏ¿´À» ¶§ °³ Ç°Á¾ Àνķü ´ëºñ ±âÁ¸ DB¿Í ¼öÁý DB¸¦ ¸ðµÎ »ç¿ëÇÑ °³ Ç°Á¾ ÀνķüÀÌ Top-1¿¡ ´ëÇؼ­ 1.5% °³¼±µÇ¾ú´Ù. °³°¡ ¾Æ´Ñ À̹ÌÁö ÀνÄÀº ·£´ý DB¸¦ 10000ÀåÀÇ °æ¿ì 93% ÀνķüÀ» È®ÀÎÇß´Ù.
¿µ¹®³»¿ë
(English Abstract)
In this paper, a dog breed image provided by Dataset of existing ImageNet[1] and Oxford-IIIT Pet Image[2] is combined with a dog breed image obtained through data mining on Internet and a random-label is added. this paper introduces to recognize 122 classes of dog breeds and 1 class that is not dog breeds. The recognition rate of dog breeds using both conventional DB and collection DB was improved 1.5% over Top-1 compared to recognition rate of dog breeds using only existing DB. The image recognition rate about non-dog image, was 93% recognition rate in case of 10000 random DBs.
Å°¿öµå(Keyword) Deep Learning   Convolutional Neural Network   Data Mining   Dog Breeds   Random-Label  
ÆÄÀÏ÷ºÎ PDF ´Ù¿î·Îµå