Site
»çÀÌÆ®
ȸ¿øÁ¤º¸
³í¹®
ÇÐȸ¼Ò°³
ÇÐȸ¼Ò°³
ȸÀå Àλ縻
ÇÐȸ¿¬Çõ
ÇÐȸÁ¤°ü ¹× ±ÔÁ¤
Æ÷»óÁ¦µµ
ÀÓ¿øÁø ¹× À§¿øȸ
¿¬±¸È¸/ÁöºÎ
À§Ä¡ ¹× ¿¬¶ôó
Çмú´ëȸ
±¹³»Çмú´ëȸ
±¹Á¦Çмú´ëȸ
ÀΰøÁö´É ¹× ·Îº¸Æ½½º ¿©¸§/°Ü¿ïÇб³
¿öÅ©¼¥
°£Ç๰
ÇÐȸÁö
·Îº¿ÇÐȸ ³í¹®Áö
Intelligent Service Robotics
ÇÐȸ¼Ò½Ä
ÇÐȸ¼Ò½Ä
´º½º·¹ÅÍ
Çà»ç¾È³»
ä¿ë°ø°í
ȸ¿ø¾È³»
ȸ¿ø°¡ÀÔ ¾È³»
Ưº°È¸¿ø
´Üüȸ¿ø
Çмú´ëȸ
Korea Robotics Society
Çмú´ëȸ
±¹³»Çмú´ëȸ
±¹Á¦Çмú´ëȸ
ÀΰøÁö´É ¹× ·Îº¸Æ½½º ¿©¸§/°Ü¿ïÇб³
¿öÅ©¼¥
Á¦19ȸ Çѱ¹·Îº¿Á¾ÇÕÇмú´ëȸ(KRoC 2024) - ÀϹÝ/Ưº°(OS)/·Îº¿ÇÐȸ³í¹®Áöµ¿½ÃÁ¦Ãâ
KRoC 2024 Çà»ç¾È³»
±âÁ¶°¿¬
Ưº°°¿¬
ÃÊû°¿¬
Ưº°¼¼¼Ç(OS)
½ÅÁø¿¬±¸ÀÚ
Flagship Conferences
ÇåÁ¤¼¼¼Ç
Æ©Å丮¾ó
¿öÅ©¼¥
¼ö»ó°á°ú
ÈÄ¿ø/Àü½Ã ¼Ò°³
µî·Ï¾È³»
> Çмú´ëȸ >
±¹³»Çмú´ëȸ
±¹³»Çмú´ëȸ
ÃÊû°¿¬
[ÃÊû°¿¬1]
ÁÖÁ¦
±ºÁý ÀÚÀ² ½Ã½ºÅÛ ÀÓ¹« °èȹ
ÀϽÃ
2024³â 2¿ù 22ÀÏ(¸ñ) 15:30~16:10
Àå¼Ò
1Ãþ Æ÷·¹½ºÆ®È¦
°¿¬ÀÚ
ÃÖÇѸ² (KAIST/±³¼ö)
¿ä¾à
º» °¿¬¿¡¼´Â ±ºÁý ÀÚÀ² ½Ã½ºÅÛÀÇ ÀÓ¹« °èȹ ±â¼úÀ» ¼Ò°³ÇÑ´Ù. ÀÓ¹« °èȹÀÇ ±â¹ÝÀÌ µÇ´Â Assignment Problem, Vehicle Routing Problem, Markov Decision Process µîÀÇ ¹®Á¦ Á¤½Äȸ¦ ¼Ò°³ÇÏ°í À̸¦ ÇØ°áÇÏ´Â ±â¹Ý À̷п¡ ´ëÇØ °£´ÜÈ÷ ¼Ò°³ÇÑ´Ù. ÀÌÈÄ, ´Ù¼ö/´ÙÁ¾ ¹«ÀÎÀ̵¿Ã¼ÀÇ Çù¾÷ ÀÓ¹«, ±ºÁý ¹«Àμö»óÁ¤ÀÇ ¹æÈ£ÀüÅõ ÀÓ¹«, °í±âµ¿ °üÃø À§¼º±º µî ´Ù¾çÇÑ ÀÚÀ² ½Ã½ºÅÛ¿¡¼ÀÇ ÀÓ¹«°èȹ ¹®Á¦¿¡ ´ëÇØ KAIST Á¤º¸¹×Á¦¾î½ÇÇè½Ç¿¡¼ ÃÖ±Ù¿¡ °³¹ßÇÑ ±â¹ýµé¿¡ ´ëÇؼ ¼Ò°³ÇÑ´Ù.
¾à·Â
- 2010. 03 ~ ÇöÀç: Á¶±³¼ö/ºÎ±³¼ö/±³¼ö, KAIST Ç×°ø¿ìÁÖ°øÇаú
- 2008. 10 ~ 2010. 02: Postdoctoral Associate, MIT Dept. Aero/Astro
- 2022. 12 ~ ÇöÀç: ÀÌÁ¾À§¼º±º ¿ìÁÖ°¨½ÃÁ¤Âû±â¼ú Ưȿ¬±¸¼¾ÅÍÀå
- 2022. 09 ~ ÇöÀç: KAIST ·Îº¸Æ½½º¿¬±¸¼ÒÀå
- 2020. 06 ~ ÇöÀç: ¹«ÀÎÀ̵¿Ã¼¿øõ±â¼ú°³¹ß»ç¾÷ ÀÚÀ²Áö´É¿¬±¸´ÜÀå
ÇзÂ
- 2004. 09 ~ 2009. 02: PhD, MIT Dept. of Aeronautics & Astronautics
- 2000. 03 ~ 2002. 02: °øÇм®»ç, KAIST Ç×°ø¿ìÁÖ°øÇаú
- 1996. 03 ~ 2000. 02: °øÇлç, KAIST Ç×°ø¿ìÁÖ°øÇаú
[ÃÊû°¿¬2]
ÁÖÁ¦
Áö»ó Çʵå·Îº¿ÀÇ ¸ðºô¸®Æ¼
ÀϽÃ
2024³â 2¿ù 22ÀÏ(¸ñ) 16:20~17:00
Àå¼Ò
1Ãþ Æ÷·¹½ºÆ®È¦
°¿¬ÀÚ
Á¶Á¤»ê (Çѱ¹»ý»ê±â¼ú¿¬±¸¿ø/¼ö¼®¿¬±¸¿ø)
¿ä¾à
½º¸¶Æ®ÆùÀÌ ¿°í ÀÖ´Â »õ·Î¿î ¼¼»ó, 4Â÷ »ê¾÷Çõ¸í ½Ã´ë, ¹Ì·¡¸¦ ¼³°èÇÏ´Â Áö±Ý, ·Îº¿ÀÇ È°¿ë ¹üÀ§µµ Á¡Á¡ ³Ð¾îÁö°í ÀÖ½À´Ï´Ù. ±×·¯³ª °øÀåÀÇ ÀÚµ¿È ±â¼úÀº ÀÌ¹Ì »ç¶÷À» ´ë½ÅÇÏ¿© ¸¹Àº ÀÏÀ» ÇÏ°í ÀÖÁö¸¸, Áö»ó ÇÊµå ºÐ¾ß¿¡¼´Â Àΰ£À» ´ë½ÅÇϱâ¶õ ½±Áö°¡ ¾Ê½À´Ï´Ù. ¸¹Àº ÀÌÀ¯°¡ ÀÖ°ÚÁö¸¸, ´Ù¾çÇÑ ÁöÇü¿¡¼ÀÇ À̵¿´É·ÂÀ» °®ÃßÁö ¸øÇÑ °ÍÀÌ °¡Àå Å« ÀÌÀ¯ÀÏ °ÍÀÔ´Ï´Ù. Àΰ£¿¡°Ô´Â ¾Æ¹«·¸Áö ¾Ê°Ô º¸ÀÌ´Â µµ·ÎÀÇ ¹æÁöÅΰú °è´Üµµ ·Îº¿¿¡°Ô´Â Å« °É¸²µ¹ÀÌ µÇ°í, ƯÈ÷ ¿¡³ÊÁö¿øÀ» ½Æ°í ´Ù³à¾ß ÇÏ´Â ÀÚ¸³Çü ½Ã½ºÅÛÀÎ º¸Çà ·Îº¿Àº È¿À²°ú ¿î¿ë½Ã°£, ±×¸®°í ±â¼úÀû ±¸ÇöÀÇ ¾î·Á¿ò µî ±â¼úÀû Á¦¾àÀÌ ´õ¿í Å©°Ô ÀÛ¿ëÇÕ´Ï´Ù. ½Ç³» ÆòÁö·Î ÀÌ·ç¾îÁø °Å½Ç¿¡¼ »ç¿ëÇÏ´Â ·Îº¿ û¼Ò±âµµ Ä«ÆêÀ̳ª ¹®ÅÎÀ» ÀÚÀ¯·Ó°Ô À̵¿ÇÏ°í Àå¾Ö¹°À» ÇÇÇÏ¸é¼ Á¦´ë·Î û¼ÒÇϱâ±îÁö 20¿© ³âÀÇ ½Ã°£ÀÌ ¼Ò¿äµÈ °ÍÀ» º¼ ¶§, ¾ßÁö¿¡¼ ·Îº¿ÀÌ ÀÚÀ¯·Ó°Ô ¿òÁ÷ÀÌ°í »ç¶÷ó·³ ÀÛ¾÷Çϱâ±îÁö ¾î´À Á¤µµÀÇ ³ë·Â°ú ½Ã°£ÀÌ ÇÊ¿äÇÒÁö °¡´ÆÇϱâ Èûµç °ÍÀÌ »ç½ÇÀÔ´Ï´Ù.
º» °¿¬Àº º¸Çà·Îº¿À» ¿¬±¸ÇÏ¸é¼ °í¹ÎÇÑ Çʵ忡¼ÀÇ ¸ðºô¸®Æ¼¿¡ ´ëÇÑ ³»¿ëÀ» °øÀ¯ÇÏ°í ÇÔ²² °í¹ÎÇϱâ À§ÇÏ¿© ÁغñÇÏ¿´½À´Ï´Ù.
¾à·Â
- 2011.02 ∼ ÇöÀç Çѱ¹»ý»ê±â¼ú¿¬±¸¿ø ¼ö¼®¿¬±¸¿ø
- 2019.02 ∼ ÇöÀç °úÇбâ¼ú¿¬ÇÕ´ëÇпø´ëÇÐ(UST) ºÎ±³¼ö
- 2011.01 ∼ ÇöÀç ±¹¹æ·Îº¿ÇÐȸ ±âȹºÎȸÀå
- 2004.08 ∼ 2006.01 Çѱ¹°úÇбâ¼ú¿¬±¸¿ø Áö´É·Îº¿¼¾ÅÍ À§ÃË¿¬±¸¿ø
- 2006.02 ∼ 2010.09 Æ÷Ç×Áö´É·Îº¿¿¬±¸¼Ò(Çö KIRO) ¼±ÀÓ¿¬±¸¿ø
ÇзÂ
- 2012~2015 ¸íÁö´ëÇб³ Àü±â°øÇаú ¹Ú»ç
- 2002~2004 ±Ý¿À°ø°ú´ëÇб³ ÀüÀÚÁ¦¾î°øÇаú ¼®»ç
- 1995~2002 ±Ý¿À°ø°ú´ëÇб³ Á¦¾î°èÃøÀü°ø Çлç
[ÃÊû°¿¬3]
ÁÖÁ¦
ÀÚÀ²¿îÇ× ¼±¹Ú ±â¼ú »ó¿ëÈ ÇöȲ
ÀϽÃ
2024³â 2¿ù 22ÀÏ(¸ñ) 17:00~17:40
Àå¼Ò
1Ãþ Æ÷·¹½ºÆ®È¦
°¿¬ÀÚ
ÀÓµµÇü (¾ÆºñÄ¿½º/´ëÇ¥ÀÌ»ç)
¿ä¾à
¼±¹ÚÀÇ ÀÚÀ²¿îÇ×Àº ÀÎÁö, ÆÇ´Ü, Á¦¾î ȯ°æÀÌ ÀÚÀ²ÁÖÇà ÀÚµ¿Â÷¿Í´Â Â÷ÀÌ°¡ ÀÖ´Ù. »ó¼± ¼¼°è 1À§ ȸ»çÀÎ HDÇö´ë ±×·ìÀº ÀÚÀ²¿îÇ× ¼±¹ÚÀ» À§ÇÑ ±â¹Ý ±â¼úÀ» ¼ö³âÀüºÎÅÍ °³¹ßÇØ¿À°í ÀÖÀ¸¸ç ±× ¼º°ú·Î ÀÚÀ²¿îÇ× Àü¹®È¸»ç ¾ÆºñÄ¿½º¸¦ ¼³¸³ÇÏ¿´°í »ó¼± ¹× ·¹Àúº¸Æ® ÀÚÀ²¿îÇ× ¼Ö·ç¼ÇÀ» »ó¿ëÈÇÏ°í ÀÖ´Ù. º» ¹ßÇ¥¿¡¼´Â ¾ÆºñÄ¿½ºÀÇ »ó¼± ¹× ·¹Àúº¸Æ® ÀÚÀ²¿îÇ× ±â¼úÀÇ »ó¿ëÈ ÇöȲÀ» ¼Ò°³ÇÑ´Ù. »ó¿ë Á¦Ç°¿¡ Àû¿ëµÈ ÄÄÇ»ÅÍ ºñÀü ¹× ¼¾¼ À¶ÇÕ±â¹ÝÀÇ »óȲÀÎÁö ±â¼ú, Ã浹ȸÇÇ ±â¼ú, °æ·Î°èȹ ±â¼ú ¹× ÀÚµ¿ µµÅ· ±â¼ú µîÀ» ¼Ò°³ÇÑ´Ù. ³·Àº ÀÚµ¿È ´Ü°èÀÇ ÀÚÀ²¿îÇ× ±â¼úÀÌ ¼±¹ÚÀÇ ¾ÈÀü, ÆíÀÇ, °æÁ¦¼º Ãø¸é¿¡¼ ÃæºÐÈ÷ °í°´ °¡Ä¡¸¦ ÁÙ ¼ö ÀÖ´Ù´Â °ÍÀ» º¸¿©ÁØ´Ù.
¾à·Â
- 2000 Çö´ëÁß°ø¾÷ µ¿¿ªÇבּ¸½Ç ¿¬±¸¿ø
- 2017 Çö´ëÁß°ø¾÷ µ¿¿ªÇבּ¸½ÇÀå
- 2018~2019 Çѱ¹Á¶¼±ÇØ¾ç µðÁöÅбâ¼ú¿¬±¸¼¾ÅÍÀå
- 2020 Çѱ¹Á¶¼±Çؾç ÀÚÀ²¿îÇ׿¬±¸½ÇÀå
- 2021~ÇöÀç ¾ÆºñÄ¿½º ´ëÇ¥ÀÌ»ç
ÇзÂ
- 2015~2022 ¼¿ï´ëÇб³ ±â°è°øÇаú(¹Ú»ç)
- 1998~2000 KAIST ±â°è°øÇаú(¼®»ç)
- 1992~1998 ¼¿ï´ëÇб³ ¹ÙÀÌ¿À½Ã½ºÅÛ°øÇÐ(Çлç)
[ÃÊû°¿¬4]
ÁÖÁ¦
¼ºñ½º ·Îº¿°ú ÇÔ²² ÇÏ´Â ¹Ì·¡ º´¿ø
ÀϽÃ
2024³â 2¿ù 23ÀÏ(±Ý) 14:10~14:50
Àå¼Ò
1Ãþ Æ÷·¹½ºÆ®È¦
°¿¬ÀÚ
À̹̿¬ (ÇѸ²´ëÇб³ ¼º½Éº´¿ø/Ä¿¸Çµå¼¾ÅÍÀå)
¿ä¾à
°í·ÉÈ·Î ÀÎÇÑ ÀÇ·á ¼ºñ½º ¹× µ¹º½ ¿ä±¸µµ Áõ°¡¿Í ÀÇ·á Àη ºÎÁ·À¸·Î ÀÎÇØ ¾î·Á¿òÀ» °Þ°í ÀÖ´Â ÀÇ·á°è¿¡¼´Â ¼ºñ½º ·Îº¿ÀÇ È°¿ë¿¡ ´ëÇÑ ±â´ë°¡ ¸Å¿ì ³ô´Ù. 7Á¾ 73´ëÀÇ ·Îº¿À¸·Î ¿¬°£ 2¸¸ °Ç ÀÌ»óÀÇ ·Îº¿ ¼ºñ½º¸¦ ¿î¿µÇÏ°í ÀÖ´Â ÇѸ²´ëÇб³ ¼º½Éº´¿øÀÇ »ç¿ë »ç·Ê¸¦ ±â¹ÝÀ¸·Î ÇöÀçÀÇ ¼ºñ½º ·Îº¿ÀÌ °¡Áö´Â ÀåÁ¡°ú ÇÑ°èÁ¡À» »ìÆ캻´Ù. ¶ÇÇÑ ÇâÈÄ ¹Ì·¡ÀÇ º´¿ø¿¡¼ ´õ¿í Àß È°¿ëµÉ ¼ö ÀÖ´Â ·Îº¿ÀÇ °³¹ß°ú È°¿ëÀ» À§ÇØ, ÀÇ·áÁø°ú ·Îº¿ Àü¹®°¡°¡ ÇÔ²² ³ë·ÂÇÒ ¼ö ÀÖ´Â ¹æ¾ÈÀ» ¸ð»öÇÏ°íÀÚ ÇÑ´Ù.
¾à·Â
- 2019~ÇöÀç ÇѸ²´ëÇб³ ¼º½Éº´¿ø Ä¿¸Çµå¼¾ÅÍÀå, ±âȹ½ÇÀå
- 2008~ÇöÀç ÇѸ²´ëÇб³ ¼º½Éº´¿ø ¹æ»ç¼±Á¾¾çÇаú ÀÓ»ó°»ç/ÀÓ»óÁ¶±³¼ö/ÀÓ»óºÎ±³¼ö/ÀÓ»ó±³¼ö
- 2003~2008 ¼¿ï´ëÇб³ º´¿ø ¼ö·ÃÀÇ, ¹æ»ç¼±Á¾¾çÇаú Àü°øÀÇ
ÇзÂ
- 2008~2013 ÇѸ²´ëÇб³ ÀÇ°ú´ëÇÐ ¹æ»ç¼±Á¾¾çÇÐ ¹Ú»ç
- 2005~2007 ¼¿ï´ëÇб³ ÀÇ°ú´ëÇÐ ¹æ»ç¼±Á¾¾çÇÐ ¼®»ç
- 1999~2003 ¼¿ï´ëÇб³ ÀÇ°ú´ëÇÐ ÀÇÇаú Çлç
[ÃÊû°¿¬5]
ÁÖÁ¦
Equivariant and Diffusion-based Methods for Data-efficient Visual Robotic Manipulation Learning
ÀϽÃ
2024³â 2¿ù 23ÀÏ(±Ý) 14:50~15:30
Àå¼Ò
1Ãþ Æ÷·¹½ºÆ®È¦
°¿¬ÀÚ
ÃÖÁ¾Àº (¿¬¼¼´ëÇб³/±³¼ö)
¿ä¾à
This talk explores innovative SE(3)-equivariant methods for end-to-end visual robotic manipulation learning, addressing challenges of data inefficiency and limited generalizability in conventional deep learning approaches. We introduce Equivariant Descriptor Fields (EDFs), highlighting their generative modeling, bi-equivariance, steerable representation, and locality principles. These enable EDFs to achieve remarkable data efficiency and generalizability without prior knowledge. We also show Diffusion-EDFs, integrating SE(3)-equivariance with diffusion-based generative modeling, enhancing data efficiency and requiring only 5 to 10 demonstrations for effective training and faster inference. Our methods demonstrate superior generalizability over previous diffusion-based techniques. We aim to uncover new research directions and solutions in the dynamic field of robotic manipulation learning through the synergy of equivariant methods and generative modeling.
¾à·Â
- 2016~ÇöÀç ¿¬¼¼´ëÇб³ ±â°è°øÇкΠºÎ±³¼ö/Á¤±³¼ö
- 2023 Ķ¸®Æ÷´Ï¾Æ ÁÖ¸³´ëÇб³ ¹öŬ¸®, ¹æ¹®±³¼ö
- 2023~ÇöÀç ¿¬¼¼´ëÇб³ ¸ðºô¸®Æ¼ ½Ã½ºÅÛ À¶ÇÕ°ú (Çö´ëÂ÷°è¾àÇаú) ±³¼ö
- 2019~2023 ¿¬¼¼´ëÇб³ ÀÚµ¿Â÷À¶ÇÕ°øÇаú (Çö´ëÂ÷°è¾àÇаú) ÁÖÀÓ ±³¼ö
- 2021~ÇöÀç ¿¬¼¼´ëÇб³ ÀΰøÁö´ÉÇаú °âÁ÷ ±³¼ö
- 2012~2016 ¹Ì½Ã°£ ÁÖ¸³´ëÇб³ ±â°è°øÇаú/Àü±âÄÄÇ»ÅÍ°øÇаú ºÎ±³¼ö
- 2006~2012 ¹Ì½Ã°£ ÁÖ¸³´ëÇб³ ±â°è°øÇаú/Àü±âÄÄÇ»ÅÍ°øÇаú Á¶±³¼ö
ÇзÂ
- 2006 Ķ¸®Æ÷´Ï¾Æ ÁÖ¸³´ëÇб³ ¹öŬ¸® ±â°è°øÇаú ¹Ú»ç
- 2002 Ķ¸®Æ÷´Ï¾Æ ÁÖ¸³´ëÇб³ ¹öŬ¸® ±â°è°øÇаú ¼®»ç
- 1998 ¿¬¼¼´ëÇб³ ±â°è¼³°èÇаú Çлç