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

»çÀÌÆ®¸Ê

Loading..

Please wait....

±¹³» ³í¹®Áö

Ȩ Ȩ > ¿¬±¸¹®Çå > ±¹³» ³í¹®Áö > Çѱ¹ÀÎÅͳÝÁ¤º¸ÇÐȸ ³í¹®Áö

Çѱ¹ÀÎÅͳÝÁ¤º¸ÇÐȸ ³í¹®Áö

Current Result Document :

ÇѱÛÁ¦¸ñ(Korean Title) Wasserstein Center ¼Õ½ÇÀ» ÀÌ¿ëÇÑ ½ºÄÉÄ¡ ±â¹Ý 3Â÷¿ø ¹°Ã¼ °Ë»ö
¿µ¹®Á¦¸ñ(English Title) Sketch-based 3D object retrieval using Wasserstein Center Loss
ÀúÀÚ(Author) Áö¸í±Ù   ÀüÁØö   ±è³²±â   Myunggeun Ji   Junchul Chun   Namgi Kim  
¿ø¹®¼ö·Ïó(Citation) VOL 19 NO. 06 PP. 0091 ~ 0099 (2018. 12)
Çѱ۳»¿ë
(Korean Abstract)
½ºÄÉÄ¡ ±â¹Ý 3Â÷¿ø ¹°Ã¼ °Ë»öÀº ´Ù¾çÇÑ 3Â÷¿ø ¹°Ã¼¸¦ »ç¶÷ÀÌ ¼ÕÀ¸·Î ±×¸° ½ºÄÉÄ¡¸¦ ÁúÀÇ(query)·Î »ç¿ëÇÏ¿© ¹°Ã¼¸¦ Æí¸®ÇÏ°Ô °Ë»ö ÇÏ´Â ¹æ¹ýÀÌ´Ù. º» ³í¹®¿¡¼­´Â ½ºÄÉÄ¡ ±â¹Ý 3Â÷¿ø ¹°Ã¼ °Ë»öÀ» À§ÇØ ½ºÄÉÄ¡ CNN(Convolutional Neural Network)°ú Wasserstein CNN ¸ðµ¨¿¡ Wasserstein Center ¼Õ½ÇÀ» Àû¿ëÇÏ¿© ¹°Ã¼ÀÇ °Ë»ö ¼º°ø·üÀ» Çâ»ó½ÃÅ°´Â »õ·Î¿î ¹æ¹ýÀ» Á¦¾ÈÇÑ´Ù. Á¦¾ÈµÈ Wasserstein Center ¼Õ½ÇÀ̶õ °¢ ¹°Ã¼ÀÇ Å¬·¡½º(category)ÀÇ Áß½ÉÀ» ÇнÀÇÏ°í, µ¿ÀÏÇÑ Å¬·¡½ºÀÇ Æ¯Â¡°ú Á᫐ °£ÀÇ Wasserstein °Å¸®°¡ ÀÛ¾ÆÁöµµ·Ï ¸¸µå´Â ¹æ¹ýÀÌ´Ù. À̸¦ À§ÇÏ¿© Á¦¾ÈµÈ 3Â÷¿ø ¹°Ã¼ °Ë»öÀº ´ÙÀ½ÀÇ ´Ü°è·Î ¼öÇàµÈ´Ù. ù ¹ø°·Î, 3Â÷¿ø ¹°Ã¼ÀÇ Æ¯Â¡Àº 3Â÷¿ø ¹°Ã¼¸¦ ¿©·¯ ¹æÇâ ¿¡¼­ ÃÔ¿µµÈ 2Â÷¿ø ¿µ»óÀÇ Æ¯Â¡À» CNNÀ» ÀÌ¿ëÇÏ¿© ÃßÃâÇÏ°í, °¢ ¿µ»ó Ư¡ÀÇ Wasserstein Áß½ÉÀ» °è»êÇÑ´Ù. µÎ ¹ø°·Î, ½ºÄÉÄ¡ÀÇ Æ¯Â¡Àº º°µµÀÇ ½ºÄÉÄ¡ CNNÀ» ÀÌ¿ëÇÏ¿© ÃßÃâÇÏ¿´´Ù. ¸¶Áö¸·À¸·Î, ÃßÃâÇÑ 3Â÷¿ø ¹°Ã¼ÀÇ Æ¯Â¡°ú ½ºÄÉÄ¡ÀÇ Æ¯Â¡À» º» ³í¹®¿¡¼­ Á¦¾ÈÇÑ Wasserstein Center ¼Õ½ÇÀ» ÀÌ¿ëÇÏ¿© ÇнÀÇÏ°í ½ºÄÉÄ¡ ±â¹ÝÀÇ 3Â÷¿ø ¹°Ã¼ °Ë»ö¿¡ Àû¿ëÇÏ¿´´Ù. º» ³í¹®¿¡¼­ Á¦¾ÈÇÑ ¹æ¹ýÀÇ ¿ì¼ö¼ºÀ» ÀÔÁõÇϱâ À§ÇÏ¿© SHREC 13°ú SHREC 14ÀÇ µÎ °¡Áö º¥Ä¡¸¶Å© µ¥ÀÌÅÍ ÁýÇÕÀ» ÀÌ¿ëÇÏ¿© Æò°¡ÇÏ¿´À¸¸ç, Á¦¾ÈµÈ ¹æ¹ýÀÌ ±âÁ¸ÀÇ ½ºÄÉÄ¡ ±â¹Ý °Ë»ö¹æ¹ýµé°ú ºñ±³ÇÏ¿© ¸ðµç ÃøÁ¤ ±âÁØ¿¡¼­ ¿ì¼öÇÑ °á°ú¸¦ ³ªÅ¸³¿À» È®ÀÎÇÒ ¼ö ÀÖ¾ú´Ù.
¿µ¹®³»¿ë
(English Abstract)
Sketch-based 3D object retrieval is a convenient way to search for various 3D data using human-drawn sketches as query. In this paper, we propose a new method of using Sketch CNN, Wasserstein CNN and Wasserstein center loss for sketch-based 3D object search. Specifically, Wasserstein center loss is a method of learning the center of each object category and reducing the Wasserstein distance between center and features of the same category. To do this, the proposed 3D object retrieval is performed as follows. Firstly, Wasserstein CNN extracts 2D images taken from various directions of 3D object using CNN, and extracts features of 3D data by computing the Wasserstein barycenters of features of each image. Secondly, the features of the sketch are extracted using a separate Sketch CNN. Finally, we learn the features of the extracted 3D object and the features of the sketch using the proposed Wasserstein center loss. In order to demonstrate the superiority of the proposed method, we evaluated two sets of benchmark data sets, SHREC 13 and SHREC 14, and the proposed method shows better performance in all conventional metrics compared to the state of the art methods.
Å°¿öµå(Keyword) ÇÕ¼º°ö ½Å°æ¸Á   ¿µ»ó °Ë»ö   µö ·¯´×   ½ºÄÉÄ¡ ±â¹Ý 3Â÷¿ø ¹°Ã¼ °Ë»ö   Convolutional Neural Network   Image retrieval   Deep Learning   Sketch-based 3D object retrieval  
ÆÄÀÏ÷ºÎ PDF ´Ù¿î·Îµå