ÀÎõ´ë ÄÄÇ»ÅͰøÇкΠÀ̼ºÈ£¡¤±è¸íÀ±¾¾ SCIE ³í¹® °øµ¿1ÀúÀÚ °ÔÀç

ÀÔ·Â:2018-12-04 20:14
°øÀ¯Çϱâ
±ÛÀÚ Å©±â Á¶Á¤
±è¸íÀ± Çлý. ÀÎõ´ë Á¦°ø

À̼ºÈ£ Çлý. ÀÎõ´ë Á¦°ø

ÀÎõ´ë ÄÄÇ»ÅͰøÇкÎ(ÄÄÇ»ÅͺñÀü¿¬±¸½Ç ¼Ò¼Ó) 4Çгâ À̼ºÈ£, ±è¸íÀ± Çлý(Áöµµ±³¼ö ¹è½Âȯ)Àº 12¿ù ¡°Learning Discriminative Appearance Models for Online Multi-Object Tracking with Appearance Discriminability Measures¡± ¶ó´Â Á¦¸ñÀÇ SCIE ³í¹®À» IEEE Access(IF:3.557, Q1) °øµ¿1ÀúÀÚ·Î °ÔÀçÇß´Ù.

À̼ºÈ£, ±è¸íÀ± ÇлýÀº ½Ç½Ã°£ ´ÙÁß °´Ã¼ ½Ã½ºÅÛ¿¡¼­ occlusion¿¡ °­ÀÎÇÑ appearance ¸ðµ¨À» ¿Â¶óÀÎ ÇнÀÇϱâ À§ÇØ subspace learning ±â¹ÝÀÇ PLS(Partial Least Square) ¹æ¹ýÀ» ÀÌ¿ëÇØ ºÐº°·Â ÀÖ´Â appearance ¸ðµ¨À» °³¹ßÇß´Ù.

¶ÇÇÑ, appearance ¸ðµ¨ÀÇ ºÐº°·ÂÀ» Æò°¡ ÇÒ ¼ö ÀÖ´Â ±â¼ú(Appearance Discriminability Measures)À» °³¹ßÇÏ¿© ÃßÀû ¼º´É ¹× ¼öÇà ¼Óµµ¸¦ °³¼±Çß´Ù.

½Ç½Ã°£ ´ÙÁß °´Ã¼ ½Ã½ºÅÛ¿¡¼­ Á¦¾ÈÇÑ ¹æ¹ýÀ» Àû¿ëÇÒ ¶§, occlusion¿¡ °­ÀÎÇÏ°í ¼º´É ¹× ¼öÇà ¼Óµµ°¡ °³¼±µÇ¾úÀ½À» È®ÀÎÇÒ ¼ö ÀÖ¾ú´Ù.

ÀÎõ=Á¤Ã¢±³ ±âÀÚ jcgyo@kmib.co.kr

GoodNews paper ¨Ï ±¹¹ÎÀϺ¸(www.kmib.co.kr), ¹«´ÜÀüÀç, ¼öÁý, Àç¹èÆ÷ ¹× AIÇнÀ ÀÌ¿ë ±ÝÁö
Ŭ¸¯! ±â»ç´Â ¾î¶°¼Ì³ª¿ä?
¸¹ÀÌ º» ±â»ç
±¹¹ÎÀϺ¸°¡ ²Ä²ÄÈ÷ Àо°í ¼±Á¤ÇÑ
¿À´ÃÀÇ Ãßõ±â»ç
±¹¹ÎÀϺ¸ ½Å¹®±¸µ¶